While there are other WebRTC solutions on the market, Subspace stands out from the rest. At Subspace, we realized the public internet was never intended for today’s real-time interactions. In seeking to solve this problem, we brought together seasoned experts from across many disciplines—distributed systems, network engineering and design, Internet peering and policy, and TCP—to create a better, more performant internet designed for WebRTC applications.

# Why the Internet Cannot Work for Real-time Applications

The internet is fundamentally broken for real-time applications. The [internet was created](https://subspace.com/resources/the-evolution-of-voice-over-the-internet) as a reliable way to transmit information. Designed as a distributed system, each network operator directly decides through Border Gateway Protocol (BGP) policy the best way to route traffic through its network. Not all operators are aligned. Some may send outside traffic over longer, higher-latency paths while routing its own traffic through better performing paths. In some cases, there is a financial benefit for the operator to route outside traffic on a less optimal path, often charging a premium for the optimal paths.

![](https://www.datocms-assets.com/41207/1644356442-lines-1.jpg)

**Internet-Grade Routers and Problems with Soft-failure Detection**

Even if you don’t consider the financial incentives that some operators might have, response times and the resolution of faults across the public internet is poor in our opinion. While modern optical networks can move traffic off a failed fiber path in milliseconds, there is a significant gap in the detection and response for soft failures. A soft failure is one in which a component in a hardware device has begun to fail but has not fully failed.  Soft failures result in an undesirable and even non-deterministic impact to the traffic traversing the affected device. A key characteristic of a soft failure is that it typically goes unnoticed until it completely fails.

In terms of complexity, internet-grade routers differ dramatically from the average home router. They usually occupy a sizable physical footprint, consume a significant amount of power, and can weigh several hundred pounds. 

![](https://www.datocms-assets.com/41207/1644007859-image5.jpg)

Examples of these internet-grade routers include:

* [Juniper Networks PTX](https://www.juniper.net/us/en/products/routers/ptx-series.html)
* [Cisco ASR 9K](https://www.cisco.com/c/en/us/products/routers/asr-9000-series-aggregation-services-routers/index.html)
* [Arista 7800](https://www.arista.com/en/products/7800r3-series)

Regardless of the manufacturer, internet-grade routers follow a similar general architecture—a chassis containing multiple line cards, one or more control plane modules, and a backplane that interconnects the line cards and control modules. Line cards are the modules where network circuits, such as copper or fiber-based Ethernet or Wavelength Division Multiplexing (WDM) connect. At a high level, the public internet is a large web of connections between devices like this all over the world.

Line cards commonly contain a number of optical transceivers, such as [Quad Small Form-factor Pluggable (QSFP) 28s](https://community.fs.com/blog/overview-of-qsfp28-lr4-optical-transceiver.html) that operate at 100 Gigabits per second (Gbps). QSFP28s use a laser to transmit data onto a fiber-optic strand.

Internet-grade routers rarely connect a single QSFP or circuit. Instead, there are hundreds of these transceivers on a single chassis spaced closely together. Each transceiver contains a laser and generates a non-trivial amount of heat. As with any electronic device, heat is a factor to a component’s lifespan, so transceivers are often one of the first components to fail.

However, the transceiver usually fails slowly over time, beginning by causing small levels of sustained packet loss and gradually increasing over time until it fails completely. This can take days, weeks, or even months.

![](https://www.datocms-assets.com/41207/1644007996-image5.jpg)

Even high-end internet-grade routers do not detect soft failures until it is far too late. In the most extreme case, the issues go unnoticed until the transceiver completely fails and the link goes down, often after months of sustained low-level packet loss.

**Detecting Failures**

[Bidirectional Forwarding Detection (BFD)](https://datatracker.ietf.org/doc/html/rfc5880) greatly improves the situation caused by soft failures, but both parties must enable BFD for each peering circuit. Even when both parties enable BFD, it still does not react until there is significant loss.

In cases where either party has not enabled BFD, severe packet loss would cause so much corruption that the routing protocol (BGP, for example) will reach a point where it is no longer able to send or receive [KEEPALIVE messages](https://datatracker.ietf.org/doc/html/rfc4271#section-4.4). The inability to send or receive KEEPALIVE messages results in the removal of that routing adjacency and routes learned from that peer.

The internet uses BGP version 4. [BGPv4 suggests](https://datatracker.ietf.org/doc/html/rfc4271#section-10) suggests a default HoldTime interval of 90 seconds and a default KeepaliveTime of one-third of the configured HoldTime. This means that one KEEPALIVE message will be sent every 30s. After 90s of not receiving one, the neighbor is deemed dead and the link logically removed from routing decisions. Several common routing vendors have adopted a more conservative default of 60/180s. On most internet-grade routers, this can be tuned as low as 5/15s.

And this is why the public internet simply cannot work for real-time applications. In WebRTC apps, 180 seconds—or even the most aggressive 15s—is simply too slow for a reaction time. By that point, traffic crossing the link has massive amounts of packet loss. Traffic crossing the faulty link may incur small amounts of packet loss for days, weeks, or even months leading up to the ultimate failure. This is the most common threat to your real-time application and one that is extremely difficult to troubleshoot and isolate. Even worse, it’s a problem on someone else's network that you have to prove exists and motivate them to correct.

# Routing and Pathfinding on Subspace’s Network

At Subspace, we understand that internet-scale routing is complex, difficult, and broken. We are experts in internet routing, networks, and in the Transmission Control Protocol (TCP). We understand that 0.1% of sustained packet loss across a given link is unacceptable for real-time computing.

![](https://www.datocms-assets.com/41207/1644007527-image3.jpg)

The Subspace approach is unique. Our global network encompasses 150+  Points-of-Presence (POP) across the world, and is one of the [top 10 most-connected Internet Exchange (IX) participants](https://bgp.he.net/report/exchanges#_participants). These IX locations are where the world’s network operators connect their routers. Our PoPs interconnect over a number of mediums, including dedicated fiber paths. While Subspace uses the BGP routing protocol at our edge to announce our Anycast prefixes, we use no standard routing protocols inside our network. Instead, traffic is routed across the best possible path using our own custom-built routing stack.

![](https://www.datocms-assets.com/41207/1644007649-screen-shot-2022-02-04-at-3-46-20-pm.png)
*Subspace network map can be found [here](https://subspace.com/network-map).*

**Real-time Precision Measurements**

To detect soft-failures, Subspace built a custom monitoring solution that sends probes between PoPs across every possible combination of ingress and egress links in real-time, all the time. This system processes more than one million measurements per second. 

![](https://www.datocms-assets.com/41207/1644007988-image4.jpg)

**Routing and Pathfinding**

While others also monitor networks to detect and alert on failures, there is a time delay between alerting on a problem and mitigating it. At Subspace, we created a better way to measure. We built a custom routing platform based on a hybrid of centralized and decentralized Pathfinding. This patented routing platform is tightly coupled with our real-time precision measurements. While others use real-time monitoring for detection alone, we use it for real-time routing decisions. This combination allows us to select the most optimal path in real-time based on real-time conditions such as latency, loss, and jitter.

One challenge we faced early on is that a purely centralized approach cannot respond in real time. With that early centralized approach, our software continuously consumed precision measurements and used that to compute a graph of the current-best paths between all PoPs. The cycle time of remotely processing precision measurements, computing the best paths, and then programming those paths back down to all PoPs was simply too long. We tuned many aspects but ultimately realized that the order of magnitude of improvement demanded by real-time computing demanded a different approach.

This led to distributed Pathfinding. With distributed Pathfinding, we run custom software on every PoP that ingests a local subset of real-time precision measurements. By using a local view of measurements, we can process them at a much faster rate.

However, immediately changing from centralized to decentralized Pathfinding does not solve all of the problems. In routing, the lowest latency loss-free path is always preferable. Any amount of sustained loss is harmful. When we observe sustained loss, we may opt for a slightly higher-latency but loss-free path. The best path may not always be direct from one PoP to another; instead the best path may traverse one or more intermediary PoPs.  

Performing routing decisions from a local view of measurements does not provide enough data.  For this reason, Subspace operates using a hybrid of centralized and decentralized pathfinding. In this distributed Pathfinding procedure, the centralized side dictates the end-to-end path while the decentralized side determines which individual link(s) to take.

![](https://www.datocms-assets.com/41207/1644440495-untitled.jpg)

**Further Refining Subspace**

But Subspace was not satisfied with “good enough.” We recently completed a significant improvement to our routing platform that will unlock the next order of magnitude of improvements.

As I explained above, reaction time with centralized pathfinding remained too slow. Unexpected PoP failures require swift reaction to avoid introducing packet loss.

To speed up centralized Pathfinding’s reaction time, Subspace changed how we process real-time precision measurements. After we introduced decentralized Pathfinding, we processed real-time precision measurements in duplicate—at both the centralized and decentralized levels.

Decentralized Pathfinding can process measurements on a much faster cycle because it only needs to consider a local view of real-time precision measurements. Instead of processing measurements in duplicate, we modified our decentralized software to report the post-calculated “best edge” to our centralized Pathfinding system and entirely removed the centralized processing.

This change unlocked that next order of magnitude of improvement. Previously, our largest bottleneck was the rate at which we could process real-time precision measurements on the centralized level. The rate at which we performed real-time precision measurements was constrained by a one-second time budget for processing those measurements at the centralized level. As our network continues to grow, we face scaling challenges while trying to remain within that one-second budget. Exceeding that budget would cause us to fall behind, resulting in Pathfinding decisions made based on past data instead of current data.

While this change dramatically improved performance, we are already developing the next major improvement to this routing design.

# Subspace Solves Pathfinding Challenges

As an engineer, I enjoy building software systems that solve difficult problems without unnecessary complexity. At Subspace, we provide a superior network experience for your real-time application. We are the world’s first dedicated network for WebRTC. With WebRTC-CDN, you can [get your real-time application running on Subspace](https://console.subspace.com/try) with a simple REST API in under 30 minutes. 

Eliminate those customer support cases related to poor network conditions. When you use the world’s only managed WebRTC service running on the premier private network created for real-time, you immediately improve your customer experience and you can focus on what matters most—building the best WebRTC application.

----------

*David Swafford is a Software Engineering Lead at Subspace and lifelong network engineer, dating back to 1997 when he installed his first home network using Cat5 cabling. Follow him on [LinkedIn](https://www.linkedin.com/in/davidswafford/) to learn more about his network engineering journey.* 

Subspace Solves Challenges of Real-time Pathfinding

[WebRTC](https://en.wikipedia.org/wiki/WebRTC) is a set of protocols that allow applications, typically running on Web browsers, to exchange media (audio, video, data) with other entities.

Before media can flow, however, the WebRTC entities need to discover what type of connection is possible, and among the possible connections, what’s the best to be used. This needs to happen as fast as possible, so that users can perceive the service as instantaneous as possible.

WebRTC includes protocols like [STUN](https://www.rfc-editor.org/rfc/rfc8489.html) and [TURN](https://www.rfc-editor.org/rfc/rfc8656.html) that are designed to facilitate the establishment of connections when a direct connection is not possible. The typical case is a computer inside a home or office network, with a private IP address, and able to reach the public Internet only through an address translation (NAT).

STUN helps in discovering the IP address and port from where a computer enters the Internet, and in some circumstances that IP address and port can be used by other entities to reach that computer. STUN is also used for keeping such bindings alive.

TURN provides a way for two entities to communicate when they are behind two different symmetric NATs, or when one is behind a firewall that restricts outbound traffic to only some UDP or TCP ports. TURN uses STUN as the underlying protocol, adding requests, responses and indications to accomplish media relay.

STUN and TURN play a role in the ICE negotiation process. [ICE](https://www.rfc-editor.org/rfc/rfc8445.html), Interactive Connectivity Establishment, is a protocol that allows the dynamic discovery of the best way to establish a connection for entities that may be behind NAT.

All WebRTC clients use ICE before media can flow.

There are three main phases: the gathering of candidates, the connectivity checks, and the nomination of the candidate pairs to be used.

The ICE candidates are simply transport addresses (IP address, port and transport type) that can potentially be used to communicate (send and receive media) and that the ICE client collects and shares to the other party through some form of signalling.

There are three main types of ICE candidates: ‘host’, ‘server reflexive’ and ‘relay’.

‘host’ candidates refer to transport addresses that are directly visible by the client, where the client can start listening for incoming connections or packets. Computers behind NAT may only have private IP addresses for ‘host’ candidates, but they are potentially usable if the other party belongs to the same network.

‘server reflexive’ candidates are the ones discovered through the interaction with a STUN server. The client sends a Binding Request to the STUN server and receives a Binding Success Response with a MAPPED-ADDRESS containing the source IP address and port from where the request was received. There may be more than one level of NAT, and that transport address represents the outmost one.

‘relay’ candidates refer to allocations reserved on a TURN server. The ICE client requests an Allocation of a relay, and after successful authentication the TURN server provides a RELAYED-ADDRESS containing the transport address allocated on that server for the client.

These types of ICE candidates have different priorities, ‘host’ being at highest priority and ‘relay’ at lowest priority; this is a way to privilege direct interconnection when possible (but that not necessarily represents the best solution in terms of connection quality and in general of Quality Of Experience).

This diagrams shows a typical process where ICE candidates are gathered and sent to the other party:

![](https://www.datocms-assets.com/41207/1647457833-ice-negotiation-1.png)

In this example the candidates are communicated as soon as they are retrieved. This technique is called "Trickle ICE" and was designed to ensure that the connectivity checks can happen as soon as the candidates are available.

Without Trickle ICE, the WebRTC client would need to wait for all the candidates to be collected before sending an offer or an answer, increasing the session set up time.

The candidates are transmitted over a signalling system established between the two parties. This is outside of the WebRTC specifications and application-specific.

The WebRTC client will then receive the ICE candidates from the other party, and it will build a list of “candidate pairs”: each local candidate will be paired with each remote candidate.

After this operation the connectivity checks can begin.

Let’s see it with an example, assuming UDP as transport for all cases. The WebRTC client has a local host candidate, IP1:port1, and has received a remote host candidate, IP2:port2. It builds a “candidate pair” with the two: {IP1:port1, IP2:port2}.

The WebRTC client will start sending STUN Binding Requests with source IP1:port1 and destination IP2:port2. These requests use STUN short-term authentication, and contain a username and password that were previously exchanged when transmitting the candidates.

If the Binding Request reaches the other party on IP2:port2, then the other party will authenticate the request, and respond with a Binding Success Response, including the MAPPED-ADDRESS attribute, containing the source of the request.

If the Binding Success Response reaches the WebRTC client, then it will identify the request by looking into the transaction ID and mark the check as Successful, and so suitable for exchanging media.

If for any reason the Binding Success Response is not received, then the candidate pair will remain in a In Progress state for some time, and then move to Failed: that pair cannot be used to exchange media.

![](https://www.datocms-assets.com/41207/1647457789-ice-negotiation-2.png)

It’s important to note that this is symmetrical: the other party too can start the connectivity check from IP2:port2 towards IP1:port1. To avoid a conflict, the parties assume the role of "controlling" or "controlled" agent. The controlling agent will be the one deciding which candidate pair will be used for exchanging media.

Before media can flow through a TURN server, a client must create a Permission. This is important during ICE connectivity checks.

For ICE candidates of type ‘relay’, the connectivity check will be performed sending Binding Requests that traverse the TURN server and reach the other party on the relay side. The Binding Requests will be carried by a Send Indication, destined to the remote candidate as peer address. The TURN server will only accept it and relay it to the destination if a Permission has been granted for that peer.

![Ice Negotiation between WebRTC and TURN server](https://www.datocms-assets.com/41207/1647457756-ice-negotiation-3.png)

(NAT has been omitted in this diagram for simplicity)

Of course a TURN allocation must exist for the CreatePermission request to succeed, but that’s already been created during the candidate gathering phase.

If the candidate pair selected for exchanging media will be one with a local ‘relay’ candidate, then typically the WebRTC client binds a TURN Channel to the other party, and starts exchanging media using ChannelData messages, instead of Send/Data Indications.

There are more details that can be discussed, like managing timeouts, role conflicts, ICE Lite, etc, which we will address in other articles. One additional aspect is important here: we mentioned three types of candidates, host, server reflexive and relayed, but there’s a fourth one, 'peer reflexive'.

Peer reflexive candidates are not provided directly by a party during candidate exchange, but are instead discovered dynamically during the connectivity checks. Getting back to the previous example with a candidate pair {IP1:port1, IP2:port2}, depending on NAT conditions, the response to the Bind Request from IP1:port1 can be received by another source, IP3:port3.

The WebRTC client can verify IP3:port3 is sending a valid response by checking the transaction ID, and if successful it will dynamically add a remote candidate of type peer reflexive. Sometimes the peer reflexive candidate is the only one suitable and will be used to exchange media.

Chrome’s chrome://webrtc-internals, Firefox’s about:webrtc and Safari's "WebRTC Logging" will show the list of candidates and the pair that was selected, so those tools are of great value when troubleshooting.

If you take a trace on the computer running the WebRTC client, you’ll be able to see the STUN Binding Requests and Responses, and CreatePermission, Send/Data Indications for connectivity checks if unencrypted TURN is used. Wireshark will filter those messages for you if you use the ‘stun’ filter, and will also be able to interpret the Binding Request/Response carried inside Send/Data Indications (and also the RTP streams, but that’s for another article).

# Subspace Provides a Managed WebRTC Network

Subspace is the only managed WebRTC network service for high-quality connectivity on the world’s only optimized, dedicated network for today’s real-time applications. 

Subspace has reimagined NAT traversal by making it an inherent part of the network rather than an add-on. We use the TURN protocol to bring traffic onto Subspace without the need for traditional TURN servers, while improving performance for **all** WebRTC traffic. With Subspace [WebRTC-CDN](https://subspace.com/product/webrtc-cdn), you get seamless and performant NAT traversal without deploying your own TURN servers.

Subspace’s network includes real-time internet weather mapping to ensure your packets are routed on the most optimal path. Our patented Dynamic Packet Routing can reduce latency by more than [80%](https://www.orange.jo/en/pages/orange-gaming.aspx) (depending on your traffic's origination and destination) and virtually eliminates jitter and packet loss.

[Try Subspace WebRTC-CDN for free now and get started in minutes](https://console.subspace.com/try).

----------

*Giacomo Vacca is a Senior Voice Engineer at Subspace and a real-time communications expert with more than 20 years of experience building and scaling VoIP, IM, and WebRTC solutions. Follow him on Twitter [@giavac](https://twitter.com/giavac).*

ICE Negotiation

[STUN](https://www.rfc-editor.org/rfc/rfc8489.html) is a protocol that helps Internet endpoints dealing with NAT (Network Address Translation), a typical condition for hosts belonging to a private network and requiring to be accessed from another network, like the public Internet.

![](https://www.datocms-assets.com/41207/1645049022-stun.png)

# Why Use STUN?

Real Time Communication services use STUN in various ways, from helping discovering the public IP address and port associated to an endpoint behind NAT (the basic “STUN service”), helping establishing connection checks for ICE ([RFC  8445](https://datatracker.ietf.org/doc/html/rfc8445)), to working as a basis for relaying with TURN ([RFC 8656](https://datatracker.ietf.org/doc/html/rfc8656)) or as a keep-alive technique for SIP OUTBOUND ([RFC 5626](https://datatracker.ietf.org/doc/html/rfc5626)).

If you'd like to learn more about TURN, then feel free to take a look at my [TURN 101 blog](https://subspace.com/resources/turn-101-subspace).

We will refer to transport addresses, which are just a combination of an IP address and a port.

STUN can be used as the basis for NAT traversal solutions, and provides three main features: 

* an extensible packet format
* coverage for multiple transport protocols
* two authentication mechanisms

In order to use STUN, a NAT traversal solution must implement a so-called “usage”, which needs to define when STUN messages are exchanged, what optional attributes to use, what type of server is used, and the authentication mechanism or mechanisms.

ICE ([RFC 8445](https://datatracker.ietf.org/doc/html/rfc8445)), SIP OUTBOUND ([RFC 5626](https://datatracker.ietf.org/doc/html/rfc5626)) and NAT Behavior Discovery ([RFC 5780](https://datatracker.ietf.org/doc/html/rfc5780)) are all examples of a STUN usage.

Furthermore STUN can be extended with additional methods, responses and attributes, like it happens with TURN ([RFC 8656](https://datatracker.ietf.org/doc/html/rfc8656)).

# How does STUN work?

STUN is a simple client/server protocol and defines two types of transactions: Request/Response and Indication (if you’re familiar with [TURN](https://subspace.com/resources/turn-101-subspace), examples are Allocate Request/Response and Send/Data Indications; indications are not reliable and cannot be authenticated.

A client sends a Request, a server sends a Response. Both client and server can send an Indication. All transactions carry a 96 bit transaction ID.

All STUN messages have a fixed header with: **Method** (e.g. Binding), **Class** (Request, Success Response, Error Response, etc) and **Transaction ID**, followed by zero or more attributes, some of which may be optional or mandatory depending on the context. Optionally, a fingerprint is added.

The header is binary, in network order, big endian, with 20 Bytes for the fixed part: 2 Bytes for the message type, 2 Bytes for the message length, 4 Bytes for the Magic Cookie (0x2112A442 - used to distinguish a STUN from other types, but also to XOR the transport addresses) and 12 Bytes for the Transaction ID.

![](https://www.datocms-assets.com/41207/1645049143-stun-message.png)

**Binding Request**

The simplest method is a Binding Request. A client sends a Binding Request to a server, and the server replies with a Binding Response containing the client’s server reflexive transport address, which is the IP address and port seen by the server as source of the request. The server reflexive transport address is returned, XOR’d, inside the XOR-MAPPED-ADDRESS attribute.

As STUN servers are typically located in the public internet, this is the easiest way for a client to discover its IP address and port on the outmost NAT.

![](https://www.datocms-assets.com/41207/1645049153-stun-binding-request.png)

**Authentication Mechanisms**

Two authentication mechanisms are defined: **short-term** and **long-term credentials**. The former is used in contexts like ICE, where endpoints exchange short-lived credentials to establish their identity during connectivity checks, while the latter is used in services like TURN, where the clients must authenticate before being allowed to create relay allocations.

With short-term credentials, authentication requires a USERNAME and MESSAGE-INTEGRITY attribute, while with long-term credentials, the client is first challenged with a Realm and Nonce attribute with a 401 Error Response (similarly to what happens with Digest Authentication in SIP), and then the client uses them with the credentials to compute the MESSAGE-INTEGRITY value leading to a successful authentication. The MESSAGE-INTEGRITY field is a HMAC-SHA1 of the message, with a key computed with just the password for short-term credentials, or with password, realm and nonce for long term credentials (again, similar to SIP).

It’s interesting to note that STUN messages can be multiplexed with other protocols, like it happens with RTP during WebRTC sessions. 

As it happens with other services, STUN servers can be located using DNS SRV lookups, where the service type is stun for unencrypted UDP or TCP (default port: 3478) or stuns for encrypted connections (TLS, default port 5349).

STUN has also a redirect mechanism (similar to the one used for SIP, and the one actually used for TURN), which is based on a 300 Error Response with an ALTERNATE-SERVER attribute containing the transport address of an alternative server where the client is expected to redirect its requests.

As we saw, STUN attributes can optionally be present, and they have a Type-Length-Value structure: 2 Bytes for the type, 2 Bytes for the length, followed by the value (padded to multiple of 4 Bytes). Standard attributes are USERNAME, ERROR-CODE, REALM, NONCE, SOFTWARE, ALTERNATE-SERVER.

![](https://www.datocms-assets.com/41207/1645042106-image3.png)

# How to Dissect STUN Messages?

Wireshark is able to detect STUN messages and to dissect them for you, so you can use the ‘stun’ filter, or more sophisticated ones like ‘stun.type.method==0x0001’ to see all Binding Requests and Responses, or ‘stun.type.method==0x0001 and stun.type.class==0x0010’ to see all Binding Responses.

Here’s an example of Wireshark dissecting a STUN Binding Request:

![](https://www.datocms-assets.com/41207/1645042270-image1-edited.png)

# Subspace Provides a Managed WebRTC Network

Subspace is the only managed WebRTC network service for high-quality connectivity on the world’s only optimized, dedicated network for today’s real-time applications. 

Subspace has reimagined NAT traversal by making it an inherent part of the network rather than an add-on. We use the TURN protocol to bring traffic onto Subspace without the need for traditional TURN servers, while improving performance for **all** WebRTC traffic. With Subspace [WebRTC-CDN](https://subspace.com/product/webrtc-cdn), you get seamless and performant NAT traversal without deploying your own TURN servers.

Subspace’s network includes real-time internet weather mapping to ensure your packets are routed on the most optimal path. Our patented Dynamic Packet Routing can reduce latency by more than [80%](https://www.orange.jo/en/pages/orange-gaming.aspx) (depending on your traffic's origination and destination) and virtually eliminates jitter and packet loss.

[Try Subspace WebRTC-CDN for free now and get started in minutes](https://console.subspace.com/try).

----------

*Giacomo Vacca is a Senior Voice Engineer at Subspace and a real-time communications expert with more than 20 years of experience building and scaling VoIP, IM, and WebRTC solutions. Follow him on Twitter [@giavac](https://twitter.com/giavac).*

STUN 101: Establishing WebRTC Sessions Around Firewalls

[Gather](https://gather.town) takes video conferencing and collaboration into the Metaverse. Gather provides a virtual layer over the real world that allows people to interact at a distance more naturally. Gather provides customizable “spaces” where participants can interact, similar to how they interact in a real space. Participants can enter and leave rooms using a UI that looks a bit like a retro video game. They can stop people they meet in the hallway for a quick chat or find someone for a quick conversation without scheduling a meeting. Compared to traditional video meeting applications, Gather Town is a whole new world.

![](https://www.datocms-assets.com/41207/1643049350-p-1-gather-metaverse-group-presentation.jpg)

# Uses for Gather

One of the problems with video conferencing is that engaging and disengaging with other participants is awkward. You have to interrupt the meeting to ask someone to speak privately and then start another session. In Gather, you can approach someone and have a private chat or make a comment to the person next to you without interrupting the larger group. 

Gather focuses on four use cases or solutions: remote work, conferences, education and social interaction. The main difference between the four solutions is the different space templates.

**Remote Work**

For [remote work](https://subspace.com/resources/what-s-next-for-the-office-the-future-of-remote-work) you can create a space that simulates your physical office, or you can create a fantasy office with moats and outdoor spaces. Each participant gets a desk that they can decorate. From your Gather office, you can easily hold a team meeting or chat with a coworker. You can share documents and work on a virtual whiteboard.

**Conferences**

You can have a main room, breakout rooms, an exhibition hall and more in a conference space. Participants can easily move from room to room or gather with a group in a corner. Presenters move up to the stage and they can share documents or their desktop. No fumbling to figure out who is presenting. Exhibitors can customize individual booth space with their own art and handouts, just as they would in a real-world conference.

![](https://www.datocms-assets.com/41207/1643049863-zxvfdsg.png)

**Education**

Teachers can create interactive classrooms. They can give a lecture to the whole class and then seamlessly divide the students into tables for small group projects. The teacher can pop from table to table to check-in and see how things are progressing.

Students also have control to create meetings, retreats, workshops for clubs or study groups, or chat with friends. There is even a provision for recess where students can take a break from their studies and play games.

![](https://www.datocms-assets.com/41207/1643050041-fg05igfxsbyoidx.jpg)

**Social Interaction**

In Gather, it is possible to host celebrations for any occasion.  You can create an event space with themed decorations and activities. Some users have created wedding venues and gotten married on Gather. It is also possible to create a watch party for games or movies. 

There is no limit to the kinds of gatherings that users can host on Gather. One of the exciting things for the Gather creators is seeing what users come up with that the creators couldn't imagine.

# How is Gather Different From Competitors?

Gather seems to have a [different mindset](https://www.gather.town/post/why-the-metaverse-matters) than its competitors. Rather than just delivering a product to do a job, like video conferencing, Gather views the problem more holistically. While humans need to see each other's faces when apart, the more significant issue is creating a human connection. Gather works to find a way for people separated by physical space to interact in a way that resembles their in-person interactions.

Gather created an online paradigm that allows humans to behave online more like they do when physically together. The way we move and engage with one another in person is different than how we behave on a conference call with pictures. By envisioning the problem differently, Gather is bridging the gap. 

![](https://www.datocms-assets.com/41207/1643050154-screenshot-2021-12-08-at-16-33-46-edited-edited.png)

There has been a lot of talk lately about the [Metaverse](https://subspace.com/resources/more-than-metaverse-subspace) and what it is. Gather is an implementation of the Metaverse that we can experience now. The Metaverse is an immersive online world where we can interact and build human connections in familiar ways to extend our experiences beyond what we can do in the physical world.

# How is Gather Driving Change Rather Than Following the Pack?

Gather hasn't invented any groundbreaking new technology to enable the Metaverse. They have taken existing technologies and combined them in new ways to produce something disruptive with a grander vision. That doesn’t mean that they haven’t introduced innovative features or that there haven’t been technical challenges to overcome. One especially interesting feature is proximity chat. In the app, you have to be virtually “close” to someone to hear them, which allows for many conversations in one space and wonderful spontaneous interactions."

![](https://www.datocms-assets.com/41207/1643049919-image-21.png)

# What Makes Gather Work Better?

A significant challenge for Gather—and anyone working on immersive, real-time applications—is the inherent limitations of the public internet. To overcome that challenge, Gather has partnered with [Subspace](https://subspace.com), the fastest, most reliable, and most secure global network available.

> "To power the kinds of meaningful, expressive, real-time interactions that are core to the Metaverse, you need reliable networking. The worst part of meeting people virtually is when you run into connection drops and lag. Unfortunately, we’ve found that the public internet lacks significantly in this regard — packet loss and general unreliability abound. Subspace fixes this with their overlay network. It allows us to make the Metaverse more accessible to people in all sorts of networking situations, all around the world."
> Kumail Jaffer, CTO at Gather

Subspace uses proprietary, AI-driven network technology to [“weather map”](https://subspace.com/resources/why-internet-weather-isn-t-going-anywhere-and-what-it-means-for-remote-work) the internet and find the fastest and most reliable path for data. By running Gather on Subspace, the things that interfere with real-time interactions and clear audio and video are virtually eliminated. 

![](https://www.datocms-assets.com/41207/1636565279-image7.jpg)

# Conclusion

Gather has produced a game-changing product and you can try it out for [free](https://gather.town/get-started). Humanity has spent a long time learning to interact in person. Gather offers a way to interact online that adapts to us rather than demanding that we adapt to the technology. Gather's innovative approach, grand vision, and next-generation networking from Subspace offer a bright future for the Metaverse—and for interacting when we can't be side-by-side. 

[Try Subspace out for free now](https://console.subspace.com/try).

How Gather is Revolutionizing Video Conferencing

Subspace, the world’s first [WebRTC delivery network](https://subspace.com/resources/accelerate-real-time-applications-on-the-world-s-fastest-network), appointed cloud industry leader Davide Petramala as Vice President of Sales.

Petramala has experience spanning three decades of leading teams and developing go-to-market strategies. He brings that expertise to Subspace, where he will oversee and lead all sales teams, develop new business strategies and opportunities, and help customers achieve maximum value with Subspace’s expanding services. 

> “We are honored to have an industry veteran like Davide join our team and bring his decades of expertise to Subspace as we continue to expand our network,” says Bayan Towfiq, CEO and Co-Founder of Subspace. “I look forward to partnering with Davide as we bring new products and services to market.”

Petramala spent the past seven years at Avaya, where he led global B2B technology sales, business development, product and channel management, and customer success. Most recently, Petramala was Managing Director of Avaya’s public cloud business unit, where he oversaw the development of the company’s cloud ecosystem of both global partners and developers while evangelizing the API economy and leading the overall market strategies for several teams. 

Petramala is an [impressive addition to the communication and networking executives](https://subspace.com/about) that comprise Subspace’s teams. Petramala, who joined Avaya through its acquisition of Esna Technologies, is excited to join the Subspace team. 

> “I’m looking forward to leading their sales team and to being in a start-up environment again.”

Subspace recently launched its network publicly, now making it possible for developers to [scale and deliver their WebRTC applications globally](https://subspace.com/product/webrtc-cdn). On Subspace, WebRTC applications are more stable, more secure, highly available, and more performant than ever before. Subspace makes it possible for companies to meet the growing demands for instant digital experiences such as metaverse environments and web3 activations. 

Petramala plans to make customers aware of those [new possibilities on Subspace](https://subspace.com/resources/subspace-is-an-internet-built-for-real-time). 

> “Going live with its global network for real-time applications in November,” he explains, “the company is now poised to accelerate its go-to-market motions and meet the increased demand for a better internet to drive amazing customer experiences.”

Want to learn more about the Subspace WebRTC solution? Check out this [white paper here](https://get.subspace.com/globalturn-white-paper/). 

[Experience the world's first WebRTC delivery network today](https://console.subspace.com/try).

Subspace Appoints Davide Petramala Vice President of Sales

In the early days of the COVID-19 lockdowns, all contact centers suddenly became remote contact centers. Around the world, contact center managers scrambled to find tech solutions to help workers keep answering calls from home. In the rush to WFH, the experience for agents and callers was not good. Distorted voices, gaps, and dropped calls are all too common. These problems are usually not the fault of the contact center solution, but rather the inherent limitations of the internet.

# How We Got Here: The Evolving Internet

The [internet is built for the reliable delivery of content](https://subspace.com/resources/the-evolution-of-voice-over-the-internet), such as email and file transfers, but not built for performance. Border Gateway Protocol (BGP)—the standardized routing protocol for the internet—has no idea what traffic it is routing or in which direction it is sending data. It simply routes data using the cheapest path possible depending on various factors, including politics, commercial agreements, and cost. As a result, certain types of data, namely real-time communications, suffer as the internet favors servicing volumetric traffic.

![](https://www.datocms-assets.com/41207/1638908569-image5.jpg)

At the start of the millennium, content delivery networks employed global networks of servers to deliver static content faster, leading the way to make OTT (Over-the-top, services offered over the top of traditional content distributors) possible with the infrastructure built by companies such as Netflix, YouTube, and Akamai. More recently, companies like Amazon Web Services have democratized cloud computing, bringing connectivity to more users in more regions of the world. 

The next evolution of the internet should be one in which the network itself is democratized, enabling speed-of-light communication for the applications that require high-velocity delivery, such as real-time communications. This evolution means delivering voice over internet protocol calls (VoIP), video and more, [regardless of the compute or data server owner](https://subspace.com/resources/voip-optimize-sdwan-sip).

# Complex Issues with a Complex Network

The internet is a complicated web of networks and billions of computers and devices that connect and work together to send and receive information. The greater the distance between the sending and receiving computers, the longer it takes the data to arrive at its destination. Information can also take different paths, which may increase or decrease the speed of delivery.

![](https://www.datocms-assets.com/41207/1642710889-image7.jpg)

A common analogy used to explain how the internet works is the postal service. In the same way that your letter travels from you to the recipient through multiple facilities (carrier, sorting, post office), data travels through numerous routers on its way to the recipient computer. Just as your letter can arrive faster depending on which mail service you use, so too can data transmit faster depending on which path it takes, though the distance between the sender and receiver has the most influence on travel time.

There are, however, a couple of significant differences in this analogy. There are generally more “stops” along the way for data. Also, the data is broken up into multiple smaller “packets” of information rather than sent as one whole mail piece.

Because there are different paths along which individual packets of information can travel, each entailing a variable number of stops along the way, packets of data can easily get lost.

Packet loss, as it’s called, results in missing information. In a VoIP call, users experience packet loss as gaps in audio. Packet loss results in strange behavior like freezing or pixelated video in video chats.

![](https://www.datocms-assets.com/41207/1638809676-image6.jpg)

Another result of these inconsistent, differing routes is variation in delivery time. Some routes are faster than others, so some packets may arrive more quickly since the internet is not conscious of which path it takes to send data from one point to another. This variation in the time it takes for data to travel across the network from one endpoint to another is called jitter. Jitter manifests as jumbled speech that creates an indecipherable conversation on a call. 

While lag (high levels of latency), jitter, and packet loss may impact any online service or application, specific applications depend upon real-time packet delivery and minimal jitter and packet loss to function correctly and maintain a reasonable quality of experience. Packet delivery, consistent latency, and minimal lag are critical for services and applications like VoIP and video chat and in numerous other industries such as online investment platforms, telemedicine, remote collaboration, virtual desktops and remote contact centers.

For these applications, milliseconds saved in delivering data can lead to a real or virtual life-and-death situation and mean the difference between quickly addressing a customer’s issue and a frustrating interaction.

# Why the Internet Doesn’t Work for Remote Contact Centers

There are four primary reasons why the internet doesn’t work for your contact center and why it will never be the answer for today’s real-time applications:

1. The cables that comprise the internet are not wired for your communications. Because a large part of the original internet’s content comes from U.S.-based entities, you can see that many undersea cables are directed to connect countries to the U.S. rather than countries to each other in a mesh pattern. As the map below shows, if your agents are in India or the Philippines, it’s likely that your Australian customers would route through the U.S. Direct paths are more in demand today but come at often unreasonably higher costs. 

![](https://www.datocms-assets.com/41207/1616444952-image8.png)
Cable map. Source: http://www.submarinecablemap.com/

2. The internet is an amalgamation of different service providers (xSPs) and Autonomous System Number/Networks (ASNs) joined together through Border Gateway Protocol (BGP) routing policies. While it is in the interest of an xSP to minimize the number of hops and cables that traffic must flow through, when it comes to ASN-to-ASN-to-ASN traffic, xSPs use BGP to facilitate traffic delivery, prioritizing the lowest cost possible through various peering and transit arrangements. That’s why your calls may not use the cable with the shortest path you see on the map above. On a less macro level, sub-optimal routing also happens inside a country and even cities. For example, it’s common to see traffic from one provider in the Philippines connect to another provider through a peering link in the United States. Though certainly not isolated, the diagram below shows a real-world example. The route from Atlanta to Washington D.C. was direct, but the return path went through California.

![](https://www.datocms-assets.com/41207/1616444937-image3.png)

3. The BGP protocol isn’t responsive to outages and congestion the way you think it should be. Unless there is a broken link, BGP doesn’t care. That means your call could see a lot of data loss and jitter, and BGP doesn’t care because the vast majority of internet traffic isn’t sensitive to either.

4. Builders of infrastructure look to where the capacity needs are—in streaming video and CDN. Service providers focus on where to get the cheapest capacity and, in many cases, actively throttle network quality. Their “best effort” connectivity guarantee might just as easily be called “least effort.” As long as it works, as long as users get their files or watch a video with a few seconds of pre-buffering, it’s good enough.

**What “Good Enough” Means for Remote Contact Centers**

The result of all those factors that make the public internet ineffective? Lag. Lag is a general term used to describe long and/or unstable latencies, i.e., jitter, and packet loss, resulting in inconsistent application behavior and poor call quality.

The latencies between your agents and your customers may be permanently slower than what is ideal and necessary. It may be so for periods that last from a few hours to several days, or calls may experience brief periods of congestion, loss, and jitter that negatively impact call quality. 

A customer experiencing lag does not enjoy the call. When calls are distorted, cut out, or drop altogether, this not only negatively affects the quality of the experience for the caller and agent, but also affects the perceived value your company places on its customers.

One of the first places a contact center is likely to see problems is in a low Mean Opinion Score (MOS). MOS, the most common metric used to measure voice and video call quality, ranges from 1 (bad) to 5 (excellent). To improve your MOS score, it is important to provide your customers with not only exceptional QoS, but also QoE. 

According to [Qualinet White Paper on Definitions of Quality of Experience (2012)](http://www.qualinet.eu/index.php?option=com_content&view=article&id=45&Itemid=52), a research project by the European Network on Quality of Experience in Multimedia Systems and Services, “Quality of Experience (QoE) is the degree of delight or annoyance of the user of an application or service. It results from the fulfillment of his or her expectations with respect to the utility and/or enjoyment of the application or service in the light of the user’s personality and current state.”

The bottom line is that the lag, latency and jitter on the public internet leads to frustrations by both the consumers and the call agents due to dropped calls, poor call quality, bad video reception, and lag. To solve this problem, the world needs a better, more simple solution.  

# How Subspace Improves Real-Time Application Delivery

A millisecond’s delay in delivery does not matter in sending an email or a text message, but it does matter in real-time communications. 
Over the past three years, Subspace has deployed custom hardware in hundreds of cities and uses its proprietary AI software to “weather map” the internet. This infrastructure gives Subspace the power to find the best paths, a combination of existing paths and Subspace’s proprietary fiber-optic backbone, and pull traffic through those paths. It is like a private carpool lane and GPS, but for dynamic internet traffic. As a result, latency and lag are reduced, and real-time applications reach their full potential.

![](https://www.datocms-assets.com/41207/1636565279-image7.jpg)

While other companies have attempted to provide solutions for expediting real-time communications on the internet, none have addressed the issue across the entire system. Many introduce other problems in their attempts. Commercial networking solutions like Cisco and Juniper were not built for global coordination or to understand and control communications traffic. Standalone solutions that require SDKs installed in clients and servers can create significant security and stability risks. And services provided by CDNs like Cloudflare and Akamai utilize the same principles and infrastructure as their primary businesses—that is, volumetric traffic—making the possibility of [optimizing real-time traffic at scale a costly and difficult endeavor](https://subspace.com/resources/voice-video-apps).

Subspace optimizes every component of network performance, from the infrastructure stack to the networking stack, including the control plane, and the data plane. With a global internet architected specifically for real-time traffic at every point, we are achieving speed-of-light communication and democratizing the network. 

Regardless of the data’s location, Subspace uses a vast system of internet quality metrics and a proprietary algorithm to direct real-time interactive traffic onto our platform and across the public and our private internet to and from servers. The algorithm can balance multiple variables to suit each application’s needs—variables such as absolute latency, any latency above a threshold, jitter and loss. The optimal delivery platform via the Subspace onramp drives better, more consistent experiences that match each application’s needs.

# Accessing Subspace

The world’s fastest, most stable, highly available and secure network is now publicly available. Subspace provides real-time application developers and companies with a quick way to operate, deploy, and scale their applications. We deliver real-time connectivity from anywhere to anywhere, and now, you can get enterprise-level connectivity anywhere in minutes, with no client-side installation required. 

Until now, Subspace has worked exclusively with large companies on their particular network needs. But now, Subspace’s WebRTC-CDN brings the capability of ubiquitous private networking to every application, developer, and internet user—expanding the possibilities for truly instant digital experiences.

Subspace WebRTC-CDN enables real-time applications to perform everywhere in the world, without client-side appliances, VPNs, or custom software. Real-time voice and video applications, multiplayer online games, modern fintech solutions, transportation apps, and every internet application in between have access within minutes to the world’s first WebRTC delivery network. You can get your contact center traffic onto our highly available, ultra-low latency, global private network with a simple configuration change.

![](https://www.datocms-assets.com/41207/1637001413-howsubspaceworks.png)

**Easy Access to Subspace: The World’s First WebRTC Delivery Network**

Subspace provides companies the capability of ubiquitous private networking for every real-time application. Developers can build instant digital experiences unencumbered by network constraints. Features of the dedicated network enterprise companies use are now accessible, self-service, by any internet application.

* **Improved quality of experience**. Faster connectivity and consistent network quality mean fewer dropped calls, more satisfied customers, happier agents, instant remote collaboration, and truly real-time experiences. On Subspace, you see a reduction in jitter, packet loss and latency by greater than 80%. With the Subspace SLA, we guarantee network accessibility and uptime. When your contact center operates on Subspace, you will always have access to the first network designed for real-time communication. When you implement WebRTC-CDN for your contact center, you can improve your MOS Score by up to 20%. 
* **Scalability and P2P Acceleration**. Unconstrained by network limitations and with company hardware management removed, companies can scale their remote worker footprint, quickly react to increases and inconsistencies in demand, and create instant digital experiences. Subspace simplifies and accelerates your P2P connections, giving your voice and video users the low-latency connections they need with a simple API call.
* **100+ Global PoPs**. With more than 100 global PoPs and growing, and high interconnection with global providers, Subspace puts your audience just milliseconds away from the global internet. No longer will it matter that your home office is in the Western U.S., your contact center is in India, and your caller is in Australia. Your audience is milliseconds away from the internet, enabling true real-time interactions.
* **Anycast Optimization Eliminates Dropped Calls and Packet Loss**. Leveraging Subspace’s proprietary AI-based anycast network analysis, Subspace routes users across the globe onto the entry point nearest their location. Continuous monitoring ensures traffic remains on the most optimal path, even when internet interference occurs. With dynamic packet route, packets travel to the destination in real-time with no packet loss. Our drop-free anycast routing means that every real-time interaction survives a network path change.
* **Higher profitability**. With improved end-user experience, customer churn decreases, more customers adopt the product, and revenue increases, while fewer requirements for company hardware and management decrease costs, altogether increasing profitability. 
* **Security**. With Subspace, you get the peace of mind that your contact center is safe and secure. You get high quality, low packet loss, with always-on, inline DDoS protection, with no additional latency. With Subspace WebRTC-CDN, we simplify enterprise-level security, giving your company, your call agents, and your customers the peace of mind they deserve.

![](https://www.datocms-assets.com/41207/1637007248-image2.jpg)
Modeled after Subspace data. Depicting packet loss from Bangalore and Miami.

Today, many of us use the internet for conferencing, and many of us experience lag during our calls—delays between callers that cause disruptions or the audio/video periodically clipping out—but accept the mediocre quality as if it's an inevitable fact of internet life. However, in customer service, millisecond delays in the delivery of data create a frustrating situation in which the interpersonal skills of the best agents cannot beat laggy internet connectivity.

# What’s Next for the Internet and Remote Contact Centers?

No, the internet was not built to support real-time communications traffic. But Subspace was. We are a dedicated network built for real-time applications. WebRTC-CDN is a managed WebRTC service that lets you quickly route your traffic over the world’s most highly available ultra low latency, global private network. With a simple configuration change, you immediately increase speed, accessibility and QoE, leading to a higher MOS score. . Your users instantly see the difference when your contact center routes traffic on the only network purpose-built for real-time applications. 

When your customer experience depends on real-time connectivity, you need Subspace WebRTC-CDN. 

[Experience the world’s first WebRTC delivery network today for free](https://console.subspace.com/try).

Why Today’s Internet Can’t Support Remote Contact Centers

[Starlink](https://www.starlink.com/) is a satellite-based internet service and a division of SpaceX. Once the Starlink “constellation” is fully deployed, SpaceX claims that every place on the surface of the earth will have access to high-speed, low latency internet service demanded by today’s real-time applications.

> "We’re really talking about something which is, in the long term, like rebuilding the internet in space."
> Elon Musk, CEO of SpaceX

The division began in 2014 and launched its first operational satellites in 2019. Ultimately, Starlink plans to have as many as 42,000 satellites encircling the globe. As of December 2021, there were more than [1,200 satellites](https://starlink.sx/) providing internet service in more than 20 countries.

![](https://www.datocms-assets.com/41207/1641401599-image5.png)

Starlink has the potential to be disruptive to the ISP industry, competing not only with other satellite internet providers such a HughesNet and Viasat, but even less performant wired technologies like DSL.

# The Challenge of Satellite Communications is Latency

Starlink’s advantage is its low latency. HughesNet, which controls about 60% of the satellite internet market, flies its satellites in Geosynchronous Earth Orbit (GEO), approximately 22,000 miles above the Earth’s surface. The long distance up to the satellite and back to the ground station introduces latency of between 500 and 800ms just to cover the distance to the satellite. This is acceptable for internet browsing and streaming video, but it is unworkable for WebRTC applications such as gaming and video conferencing. 

The impact of such long latency is obvious when watching a news broadcast with a remote reporter. When a reporter is in the field and linked to the studio via satellite, there is a noticeable pause between the time that an anchor asks a question and the reporter answers. That pause is the lag introduced by routing the communication through a satellite in GEO. 

![](https://www.datocms-assets.com/41207/1641401580-image3.png)

Starlink, by comparison, places its satellites in Low Earth Orbit (LEO) at 340 to 750 miles above the surface. The shorter distance adds only 10ms of latency, which is a lag easily tolerable by most real-time applications. This could be a game changer for rural residents who don’t have access to wired broadband and provide even more options for [home workers](https://subspace.com/resources/remote-work).

# In Space Bandwidth is Also a Limiting Factor

Bandwidth is usually not a problem for wired broadband communications. Zoom video conferencing only requires around 3Mbps for HD video calls. Many interactive competitive games require about the same. For example, Fortnite requires 3Mbps while Call of Duty requires 4Mbps. HughesNet offers 25Mbps and Viasat has plans that go to 100Mbps. While these can’t compete with Gigabit fiber connections, 25Mbps should be adequate for streaming a movie and attending a Zoom meeting at the same time. 

The bandwidth problem comes in when users exceed their monthly allotments. If a user exceeds their monthly bandwidth limits, both HughesNet and Viasat drastically throttle the speed, down to as little as 1Mbps for the remainder of the month. That’s not enough for a Zoom call, let alone enough to do anything else at the same time. This throttling happens because HughesNet and Viasat have a limited number of satellites to service all their users. Viasat only has one satellite covering all of its subscribers in the US while HughesNet only has two. 

During Starlink’s beta test, they had approximately 70 satellites connected to 50 ground stations to serve the US and that number is projected to increase by tenfold by 2026. By using many smaller satellites, Starlink is able to offer more usable bandwidth to each customer with no usage limits. In addition, Starlink’s flexible constellation can be increased to meet demand as it grows. [Speedtest.net reports](https://www.speedtest.net/insights/blog/starlink-hughesnet-viasat-performance-q2-2021/) that the actual download speed for Starlink is 97.23Mbps compared to 19.73Mbps and 18.13Mbps for HughesNet and Viasat respectively. 

![](https://www.datocms-assets.com/41207/1641401591-image4.png)

# How Does Starlink Work?

Starlink is a constellation of satellites in Low Earth Orbit (LEO), between 340 and 750 miles above the surface. The satellites connect to a consumer’s home network through a satellite dish, a little smaller than a typical satellite TV dish, and a Starlink modem. 

The satellites communicate with the ground using the [-Ka and -Ku radio bands](http://www.aidforum.org/topics/technology-data/ka-vs-ku-band-which-is-the-best-for-satellite-broadband/). The shorter frequency -Ka band is faster while -Ku is better at penetrating weather and objects that may obstruct the line of sight between the satellite and the ground. Starlink uses the -Ku band to communicate with user terminals and the faster -Ka band to connect with ground stations. 

![](https://www.datocms-assets.com/41207/1641402586-dsc06258.jpeg)

The satellites communicate with one another via laser. The consumer’s message is passed from satellite to satellite until the message reaches the nearest ground station. The message is then passed to the nearest Starlink Point of Presence (PoP) on the terrestrial internet where it is handled like any other internet traffic. 

Starlink has an [agreement](https://www.theverge.com/2021/5/13/22433982/elon-musk-spacex-internet-connectivity-deal-google-cloud) to have a ground station and PoP at every Google data center, closely tying Starlink to the terrestrial internet. Laser-equipped satellites can also be used to transmit data over long distances. A laser in the vacuum of space can transmit data up to six times faster than fiber optic cable. Laser-equipped satellites in LEO could substantially outperform undersea fiber optic cable for trans-oceanic communications.

> "Laser links in orbit can reduce long-distance latency by as much as 50%, due to higher speed of light in vacuum and shorter path than undersea fiber."
> [Elon Musk, CEO of SpaceX](https://twitter.com/elonmusk/status/1415527753185730564) 

Each satellite in the constellation is about the size of a large dining table—about nine feet long and weighing 560 pounds. The surface of the solar panel is substantially larger than the satellite body at around thirty feet long and nine feet wide. Each satellite is equipped with a [Krypton Hall effect thruster](https://www.sciencedaily.com/releases/2011/10/111006084023.htm). 

The thruster uses electricity to ionize krypton molecules and then directs a plasma stream using a magnetic field to create thrust. The thruster is used to maintain altitude, maneuver the satellite and deorbit at the end of the satellite’s life. The Hall thruster is so efficient that the satellite’s krypton fuel will last for years. The satellites are launched sixty at a time onboard the SpaceX Falcon 9 rocket.

![](https://www.datocms-assets.com/41207/1641401620-image7.png)

# How is Starlink Different From Other Satellite Internet Companies?

Starlink has competitors in the consumer satellite internet business. In addition to HughesNet and Viasat, which have been offering satellite-based internet for decades, a new competitor—[Project Kuiper](https://www.aboutamazon.com/news/innovation-at-amazon/project-kuiper-announces-plans-and-launch-provider-for-prototype-satellites)—is trying to catch up with Starlink. Kuiper is an Amazon initiative. It plans to launch two prototype satellites in late 2022. That makes Kuiper more than three years behind Starlink. Kuiper plans to start service once it has launched 578 satellites and to eventually have 3,236 satellites in its constellation.

In November 2021, aerospace giant [Boeing entered the satellite internet business](https://www.theverge.com/2021/11/3/22761963/boeing-satellite-internet-network-constellation-spacex-amazon) with FCC approval of its 147-satellite constellation offering broadband internet in the U.S., Puerto Rico, and the U.S. Virgin Islands. Boeing has not released a schedule, but its FCC permit requires the company to launch half of its satellites before 2028. Boeing satellites will operate in the V-Band, which is a higher frequency range than either -Ka or -Ku bands used by Starlink, potentially providing more bandwidth but at the cost of more interference. 

# What Does This Mean for the Future of the Internet?

LEO satellites, such as Starlink, will revolutionize high-speed internet for those outside of major cities. Where wired broadband internet is not available, consumers’ choices are limited. Currently, the most popular option in the US is HughesNet, which adds nearly a second of lag, far more than most WebRTC applications can tolerate.

Starlink seems likely positioned to avoid a key problem that has plagued early offerings of internet and mobile phone technology. For internet service, DSL was the first broadband infrastructure brought to a wide consumer market. DSL was the leader in an emerging market. As other, faster options started to become available, DSL providers found themselves saddled with an expensive investment in an infrastructure that was difficult to upgrade. 

The same thing has happened in successive waves of cell phone technology. Soon after one early adopter invested in 2G, the next provider was offering its customers 3G, then another had 4G. 

Starlink satellites have a five-year projected lifespan and the design of new satellites being launched is improved with each successive deployment. This strategy means that by the time Starlink’s next competitor launches its first generation satellite, Starlink will be on its third or fourth generation and the entire fleet will be renewed every five years. That ensures the Starlink service will always run using the latest technology.

Starlink has big ambitions. Its current user contract includes a [Mars Provision](https://www.inverse.com/innovation/spacex-mars-city-starlink-terms-of-service-declares-free-planet). No other internet service provider has a contract clause for Mars.

> "For Services provided on Mars, or in transit to Mars via Starship or other colonization spacecraft, the parties recognize Mars as a free planet and that no Earth-based government has authority or sovereignty over Martian activities. Accordingly, disputes will be settled through self-governing principles, established in good faith, at the time of Martian settlement."
> From Starlink consumer contract, as reported by [UniverseToday.com](https://www.universetoday.com/148733/one-of-the-terms-of-service-for-starlink-is-that-you-recognize-mars-as-a-free-planet/)

Starlink fills an important gap in internet service by going were cables cannot and do not exist. It also has a strategy that has brought a solution to market first and could future-proof it against competitors. This all converges with a mass social trend that [disperses workers away from centralized offices](https://subspace.com/resources/what-s-next-for-the-office-the-future-of-remote-work) and increases the demand for high-speed internet in more and more places. It is an exciting time for the internet and maybe bringing bigger changes than the dot com boom of te early 2000s.

# Starlink is Primarily a Last Mile Solution

The primary purpose of Starlink is to link a business or household to the internet via a low orbiting satellite. This is revolutionary technology that will expand high-speed internet to millions of underserved customers all over the world, but it is a [last mile solution](https://subspace.com/solutions/address-the-last-mile). Once the user’s message reaches the ground station, it is still at the mercy of middle mile problems that all public internet users experience—latency, jitter, and packet loss. The public internet [just wasn’t designed for today’s real-time applications](https://subspace.com/resources/the-evolution-of-voice-over-the-internet) such as video conferencing, VoIP, and gaming.

[Real-time application developers and publishers](https://subspace.com/resources/accelerate-real-time-applications-on-the-world-s-fastest-network) who want their Starlink-enabled customers to get the full benefit of their applications should run their applications in [Subspace](subspace.com).

Subspace’s groundbreaking real-time network infrastructure and services platform provide the lowest latency, most reliable, real-time, and fully controllable network possible for the world’s biggest applications. On Subspace, WebRTC applications see an expansion in usable latency, larger user pools, reduced user churn, and broader geographic reach—leading to a steep uptick in engagement across the applications’ ecosystem. Subspace has solved the middle mile problem. In addition, Subspace’s simple to use [GlobalTURN](https://subspace.com/product/globalturn) product allows developers to put their WebRTC application on Subspace [within minutes](https://subspace.com/resources/globalturn-quickstart-guide). GlobalTURN also drastically reduces maintenance costs over a homegrown TURN solution. [Try Subspace now for free](https://console.subspace.com/try).

SpaceX is Giving the Internet Lift With Starlink

# Networks, Network Performance, and You

Whether your customers are playing a video game with their friends from across the world, talking with their relatives in another country, or working from home, anything that goes wrong with the network between the customer and what’s important to them is a cause for concern. If your customers can’t make the network perform, your application is simply not going to work. 

But Networks are complicated.  Between your customer and the service they want to access sits miles and miles of cables and wireless distance, routers and switches, and network exchanges. Anything going wrong can create all sorts of connection problems, making your desired service unusable. Being able to troubleshoot network performance and understand the issues is essential. How do you measure network performance so that you can fix the issue?

![](https://www.datocms-assets.com/41207/1638908569-image5.jpg)

# Key Network Performance Metrics to Measure

When your customers complain about freezing video or slow download speeds, you need to be able to troubleshoot the network. There are several key network performance metrics that you need to examine. 

**Latency**

Information sent through the network takes time to go from the user to the data center. This time is called “latency”, and it measures the number of milliseconds it takes from when a packet is sent from you to when your service receives the packet. here’s also the return time it takes for the service to send a packet to you. The  Round Trip Time (RTT) is the sum of both of those.

![](https://www.datocms-assets.com/41207/1636565279-image7.jpg)

In the networking world, time and distance correlate tightly. So a long RTT means that you are physically very far away from the service you want to use, and a short RTT implies that you are physically close to the service.  

The closer you are to the service, the less likely that something will go wrong. The further you are, the greater the likelihood of errors. When you’re talking halfway around the world, like New York to Delhi in India, the only question is how much is going wrong, not if something will go wrong. 

**Jitter**

After latency, the next question is regularity. If the first packet you send out takes 100ms to reach your desired service, will the second one also take 100ms?  Or will it take longer, like 120ms?  If every packet takes the exact same amount of time between being sent and received, the jitter is 0.  If every packet takes a different amount of time, there is jitter.  If every packet takes a different amount of time, and that time difference is really big, then there is a lot of jitter.

**Packet Loss**

![](https://www.datocms-assets.com/41207/1638809676-image6.jpg)

One cause of jitter is packet loss.  If you send out one packet to your desired service and it doesn’t arrive, your device has to wait to figure out that it’s timed out, and must retransmit that packet.  So from the service’s perspective, it thinks things are fine, since it received one packet, but from your device, it knows it sent two, and all of a sudden there’s a long time between when it first tried to talk with the service and the service responded, so it calls this variability in time jitter.

**Throughput (aka “Bandwidth”)**

When there is jitter and packet loss, then throughput goes down.  Throughput is just the amount of data which can be sent per second.  If every other packet is lost, then observed throughput is just half of what it would be if no packet was lost.

# How do you know the state of your network?

You have to measure.  Fortunately, there are a lot of available network performance tools available which are free to use, in common use, and will tell you an awful lot about what’s going on.  Each tool tells you something a little different than the others, and each one specializes in a certain aspect of overall network performance.

* **[iPerf](https://openmaniak.com/iperf.php)**
	* iPerf has to be run on both a device and a server, and it measures throughput between them. 
* **[Speedtest by Ookla](https://www.speedtest.net/)**
	* Speedtest will show you your observed throughput between Ookla’s servers and your device.  This is the bandwidth your device actually sees, so it’s net of all possible loss and jitter your network may have.
* **[Packet Loss Test](https://packetlosstest.com/)**
	* Packetlosstest.com tells you your device’s packet loss, latency, and jitter from within your web browser.  If your service isn’t accessible from a web browser, you can still use it to see a part of your network, even if not the whole of it.  And again, this is between Packetlosstest’s servers and your device.
* **[MTR](https://www.tecmint.com/mtr-a-network-diagnostic-tool-for-linux/)**
	* MTR is a wonderful tool which shows the network path and status between your device and desired service, but it’s very technically oriented and lacks the sort of user interface the other tools offer.    	
* **[Netperf](https://hewlettpackard.github.io/netperf/ )**
	* An oldie-but-a-goodie.  Netperf consists of a client and server, you run the client on your device, usually a computer, and the server on the server you’re trying to talk to.  It takes some expertise to use this tool, but it’s worth the effort.

But what can you do about it all?  If you start to measure and monitor and discover that there’s problems with packet loss, jitter, or latency, what is there to be done?

Use Subspace.  

Subspace has designed and built a Network-as-a-Service, you can think of it as a parallel Internet which is faster and more reliable than the general Internet.  This means not just lower latencies than the Internet, it also means reduced jitter and packet loss, so it’s a global network which simply works better.

![](https://www.datocms-assets.com/41207/1637005966-t-1.jpg)

# How can you use it?

Let’s take our tool iPerf. We’ll bring up iPerf on a server in Amazon’s cloud.  Then we’ll bring up an iPerf client on a local laptop.  Then the secret sauce:  we’ll create a Subspace [PacketAccelerator](https://subspace.com/product/packetaccelerator) between the laptop and the server.  And we’ll know by how much Subspace was faster by running iPerf first over the Internet, and then over the Subspace PacketAccelerator.

First, let’s create the PacketAccelerator ([sign up for free here](https://console.subspace.com/try)):

![](https://www.datocms-assets.com/41207/1641245943-image2.png)

Then let’s start up iPerf on the server:

![](https://www.datocms-assets.com/41207/1641245991-image4.png)

With iPerf on the server, let’s start the iPerf client on the laptop, and connect directly over the Internet to the iPerf server:

![](https://www.datocms-assets.com/41207/1641245976-image3.png)

It completes, and returns an observed bandwidth of 11.6 Mbps.

Now let’s run the iPerf client on the same laptop, but direct it to use the Subspace PacketAccelerator instead:

![](https://www.datocms-assets.com/41207/1641245938-image1.png)

And sure enough, the PacketAccelerator is faster, achieving 16.3 Mbps instead of the 11.6 achieved over the public Internet.

# Wait, so what is Subspace?

[Subspace](https://subspace.com) provides real-time application publishers with a platform to operate, deploy, and scale their software as a network platform as a service. Our groundbreaking real-time network infrastructure and services platform provide the lowest latency, most reliable real-time, and fully controllable network possible for the world’s biggest applications. On Subspace, real-time applications see an expansion in usable latency, larger user pools, reduced user churn, and broader geographic reach—leading to a steep uptick in engagement across the application’s ecosystem.  

Subspace delivers the experience users deserve—across the world and across all titles and platforms. [Try Subspace out for free now](https://console.subspace.com/try).

How to Measure Network Performance

**TL;DR**

Not only is jitter frustrating for call quality, but it can also lead to customer dissatisfaction and revenue loss. How can it be measured and reduced? Find out in this VoIP jitter guide. Subspace, because it’s designed differently, is far better suited to these more demanding and emerging applications. Decide for yourself by [trying Subspace for free now](https://console.subspace.com/try).

Estimated read time: 7 minutes

------------------------------

> “– is – the most – things that can happen – voice call.”
 
Let’s try that sentence again.

Jitter is one of the most frustrating things that can happen on a voice call.

Also sometimes called stuttering or ping spikes, jitter occurs when packets of data that flow between parties on a VoIP call don’t reach their destination as expected.

There’s lots of information that gets exchanged between devices over the course of a voice call. This data is divided into small pieces of information, known as packets, which are transmitted over the internet in a continuous stream, each taking their own path. When these packets reach their destination—the other device involved in the call—they are reassembled in order and converted into an audio signal the receiving party can hear via a digital signal processor (DSP).

While this may sound pretty straightforward, at any given time, there are millions of packets of data all flowing across the internet, trying to reach their destinations. This means that the journey doesn’t always go as planned. Some packets may be delayed and arrive out of order, and some may never reach their destination at all.

![](https://www.datocms-assets.com/41207/1640190702-image4.gif)

To fix jitter, you’ll need to improve how packets move between the VoIP call’s endpoints.

# What Are the Overall Impacts of Jitter?

If only a small amount of data is lost or delayed during a VoIP session, it’s generally not a problem. The DSP can interpolate what was lost and fill in the missing data, with no noticeable impact on the receiving end. But if packet loss or delays exceed the DSP’s capabilities, it will affect call quality.

The uneven transmission of data packets is technically known as packet delay variation (PDV), but a user experiences it as jitter, stuttering, or ping spikes.

What does jitter sound like? Unfortunately, just about anyone who has ever made a VoIP call probably knows all too well what it sounds like: dropped words, jumbled sentences, and other annoying audio problems.

Anyone who has ever tried to have a conversation while experiencing a high level of jitter knows how frustrating this problem can be.
 
With more people working from home, including remote call center reps, jitter is more than just an annoyance. For example, if businesses use VoIP to support phone interactions with customers, calls plagued with jitter can erode customer satisfaction and phone agents’ ability to effectively close sales and resolve customer queries. (Don’t think that’s a big deal? Today, 96% of consumers will consider switching to another business after a [bad customer experience](https://www.forbes.com/sites/shephyken/2020/07/12/ninety-six-percent-of-customers-will-leave-you-for-bad-customer-service/#open-in-new-tab).) Adding to the business risk, jitter can also make it difficult for employees to work together and collaborate remotely.

![](https://www.datocms-assets.com/41207/1640190698-image3.jpg)

While the occasional lost or delayed packet won’t likely cause disruption, there’s a threshold at which jitter becomes noticeable. As a rule of thumb, less than 1% packet loss won’t cause dropped words or other call quality problems. Jitter is also less likely to be an issue as long as individual packet delivery delays are less than 30 milliseconds (according to Cisco), and overall network latency is less than 150 milliseconds.

# How to Measure and Test Jitter

Because of the potential impacts of jitter, it’s a good idea to understand what’s going on with your voice apps.

Jitter can be measured using the average packet-to-packet delay time. This is done using proprietary or open-source software that analyzes various quality of service and VoIP call metrics that identify potential contributors to jitter, such as network congestion and packet loss. Some examples include [SolarWinds Network Jitter Monitoring](https://www.solarwinds.com/voip-network-quality-manager/use-cases/network-jitter/#open-in-new-tab) and the open-source [StarTrinity Jitter and Packet Loss Test Tool](http://startrinity.com/InternetQuality/JitterPacketLossTest.aspx/#open-in-new-tab).
 
![](https://www.datocms-assets.com/41207/1640190706-image6.jpg)

The analytics provided by some networking products, such as Cisco IOS voice gateways and routers, can also be used to diagnose excessive jitter.
 
But regardless of what tool you use to test jitter, your ability to fully illuminate what’s happening on your VoIP calls depends on whether you have control over one or both endpoints that are involved.

**One endpoint**. If you only have access to performance data from one endpoint, you’ll have to conduct a ping jitter test to measure and work out both the mean and minimum round-trip times for a series of packets (again, keeping in mind the thresholds outlined above at which jitter becomes problematic). Round-trip time (RTT) is the amount of time it takes a packet to travel from the source to its destination and back again.

**Two endpoints**. With two endpoints, you’ll have greater visibility into how packets are exchanged. In this case, you’ll use instantaneous jitter measurement, which is the variation between transmission and receipt of each packet. This can then be used to determine average instantaneous jitter on a call.

# How to Reduce Jitter for VoIP

If you’ve ever experienced jitter, then we probably don’t need to tell you it makes for a bad calling experience. And you’re probably wondering how to improve latency and consistently reduce jitter on all your calls.

There are a number of tactics that are helpful for reducing jitter on VoIP calls.

**Optimize your bandwidth**. Think of how much faster you can reach your destination on a road trip when the highways are clear. Similarly, you can reduce latency—and jitter—by deploying practices that reduce traffic by limiting network usage and reducing bandwidth hogs. These may include:

* Blocking or limiting streaming sites
* Throttling cloud backup applications
* Using hosted filtering
* Centralizing application/OS updates

**Use an ISP that provides enterprise-grade VoIP services**. Many consumer-grade ISPs aim to deliver a great web surfing experience. But while they may all generally support the protocols involved in making VoIP calls, they don’t necessarily prioritize this type of traffic. A business-class high-speed solution may provide calling that’s more consistently jitter-free.

**Use a router that supports packet prioritization**. It may also be that your router isn’t up to the task of providing low-latency, jitter-free calling. This can be improved with a router that offers packet prioritization, so that the data associated with voice calls is given priority—and thus reaches its destination faster—than other types of traffic.

![](https://www.datocms-assets.com/41207/1640190711-image7.jpg)

**Deploy a jitter buffering device**. Another way to keep calls flowing is by installing a jitter buffer on your VoIP system. While a jitter buffer doesn’t help to improve latency in transit (in fact, it slows VoIP traffic down), it improves call quality by holding packets until they arrive, putting them back in order, and retransmitting them in the correct sequence. It’s important to note that a buffer won’t fix the cause of jitter (as the above tactics can). But it can help to provide clearer calls with fewer interruptions.

**Adjust your hardware**. There are a couple of quick hardware adjustments that can help to improve the quality of your calls. For example, running IP phone devices above the 5 GHz range may cause network interference. Ensuring phones are running in the 2.4 GHz range can help. In addition, newer ethernet cables (Cat 6 and above) are capable of transmitting data at much higher speeds, improving VoIP performance.

# Enter Subspace

At the end of the day, reducing jitter is all about speed—and for the best-quality call experience, you’ll need a network that’s designed to meet the needs of VoIP.
Subspace’s network runs parallel to the Internet, giving companies a high-speed alternative for their VoIP services. This means low latency, reduced jitter, and less dropped packets.

[Try Subspace now for free](https://console.subspace.com/try).

VoIP Jitter: Guide for Testing, Measuring and Reduction

Dr. Mark van Rijmenam, [“The Digital Speaker,”](https://www.thedigitalspeaker.com/) works with the world’s top companies to use emerging technologies to create digital transformation. As Van Rijmenam explains, “Digital transformation is more than getting online. Companies must be digital.” In a recent conversation, Van Rijmenam explained why digital transformation is essential, why it must happen now for companies to succeed, and how the current internet infrastructure must change to accommodate those demands. 

![](https://www.datocms-assets.com/41207/1640187957-headshot-dr-mark-van-rijmenam.jpg)


# The Digital Speaker Joins the Metaverse

Dr. van Rijmenam describes himself as “a future tech strategist, international keynote speaker, author and expert on disruptive innovation and how we can benefit from emerging technologies such as big data, blockchain, AI and VR/AR.” 

He began his career in hospitality management and has gone on to receive a Ph.D. in Management from the University of Technology Sydney. He is the founder of [Datafloq](https://datafloq.com/) and co-founder of [AKVO](https://akvo.io/). He recently launched the [Between Two Bots podcast with Dan Turchin](https://betweentwobots.com/), covering the cultural, technological and ethical implications of artificial intelligence, technology and the metaverse. 

Van Rijmenam traveled the world pre-2020, speaking and delivering his message. He tells companies that “digital transformation is more than getting online—companies must be digital.” 

That became more evident than ever with the 2020 pandemic. Each December, Van Rijmenam makes annual trend predictions for the coming year. He remembers back on the projections he released in December 2019. 

“If I had said that within three months, 95% of the world would work remotely, everyone would have called me an idiot. But that’s what happened. The pandemic kick-started the digital age. It pushed us into the digital age, and we didn’t have a choice.” 

As a result of the pandemic—and taking his own advice to be digital—Van Rijmenam created The Digital Speaker, a concept where his avatar speaks virtually to real-world audiences. In fact, on October 10, 2020, he delivered the world’s first TEDx Talk in virtual reality, discussing the rise of digitalism. Today, he speaks in person, virtually, or as an avatar or hologram in what might be considered the metaverse.  

![](https://www.datocms-assets.com/41207/1640030239-image4.jpg)

# The Metaverse: Are We There Yet?

Van Rijmenam explains that the metaverse is the “immersive internet, the next version of the internet.” 

But are we there yet, and when will we know when the metaverse is here? 

“Some say we are already there,” explains Van Rijmenam. “It’s much like mobile and internet 2.0—when did it start? Was it with the introduction of the iPhone or was it before that? If you see what’s happening with Roblox, Minecraft, Fortnite, Meta and work on VR at Microsoft, then yes, we are already merging the physical and digital worlds.”

There is now even a term for this merging of physical and digital worlds—phygital. In a [June 2021 Forbes article](https://www.forbes.com/sites/forbesbusinesscouncil/2021/06/30/phygital---what-is-it-and-why-should-i-care/), strategist Paul Prior explains that phygital is “the combination of physical and digital for enhanced experiences.” 

![](https://www.datocms-assets.com/41207/1640030279-image7.jpg)

He points out that society is already experiencing a bit of this phygital world through companies like Amazon, Alibaba, Nordea and b8ta. He uses the Amazon Go Store as an example, where shoppers explore the store, and items immediately charge to their credit card when they depart the physical location.  

![](https://www.datocms-assets.com/41207/1640030227-image3.jpg)

In a [recent interview for Morning Brew, Craig Donato, chief business officer at Roblox](https://www.morningbrew.com/emerging-tech/stories/2021/12/10/roblox-s-metaverse-is-already-here-and-it-s-wildly-popular?), explained that things are moving toward the metaverse, but it remains a work in progress. While the platform has 47 million daily active users, 9.5 million developers and more than 24 million player experiences, Donato emphasizes that the world is only in the infancy of the metaverse. “We are still in the very early days of the metaverse, and it will be a while before we see things like interoperability between metaverses and full immersion.” 

Van Rijmenam agrees with Donato. “Yes, we are already merging the physical and digital worlds, but the Metaverse—the magical space where digital and physical worlds meet—isn’t here yet. It will take a while to make that happen. The interoperability, openness, social components aren’t there yet.”

# Why Does a Digital Transformation and the Metaverse Matter to Business?

Van Rijmenam explains that we were on the edge of the digital age, but the pandemic propelled society forward in 2020. “We must be digital. We stand at the start of an exponential decade and it’s about to explode.” 

Van Rijmenam points to the youngest generations, [Gen Z and Gen Alpha](https://mccrindle.com.au/insights/blogarchive/gen-z-and-gen-alpha-infographic-update/). Generation Z (those born between 1995 and 2009) and even the youngest members of society, Generation Alpha (those born since 2010) will demand a digital experience. 

“To reach Gen Z or Gen Alpha, we see that they are metaverse natives. They live in Roblox. They see their friends more often in virtual worlds than in real worlds.” 

![](https://www.datocms-assets.com/41207/1640030318-image10.jpg)

He reminds us that the earliest Gen Z members are already in the workforce while the oldest Gen Alphas will reach the workforce in the next 10 years. “They will expect a different approach. Technology will not stop. Organizations need to be aware of how they can benefit from the next version of the internet.” 

![](https://www.datocms-assets.com/41207/1640030209-image1.jpg)

If company leaders wait until the world around them is digital, it is too late. “If you want to remain relevant in the next decade, you need a digital transformation,” explains Van Rijmenam. But to enable that transformation, Van Rijmenam explains that we must solve the [problems inherent in today’s public internet](https://subspace.com/resources/the-evolution-of-voice-over-the-internet). 

# Digital Transformation Requires a New Infrastructure

Despite this apparent need for an immersive experience and a digital transformation, Van Rijmenam cautions that we have much work to do. He points out that today’s internet was developed decades ago and never intended for immersive experiences. 

![](https://www.datocms-assets.com/41207/1640030216-image2.jpg)

“What is important [to immersive internet] is the compute needed. It’s enormous.” He explains that latency, lag and jitter aren’t something you can tolerate when you are in an immersive experience. “We must have the infrastructure in place to make that happen.”

Van Rijmenam can found on Twitter [@vanrijmenam](https://twitter.com/vanrijmenam).

That is where [Subspace can help your company](https://console.subspace.com/try) meet the demand for digital transformation. To indeed be digital, you need to operate in real-time, without worrying about frozen video, lag, and old connections. You need a new internet designed for today’s digital world. Subspace is that solution. 

![](https://www.datocms-assets.com/41207/1639497622-image5.png)

[Subspace](https://www.youtube.com/watch?v=2GZV-wZvKSo) is the fastest, most stable, highly available and secure network for internet applications. 

Subspace deploys, operates, and scales high-performing real-time, global networking-as-a-service for voice and video applications and WebRTC applications. Subspace disrupts the public internet, paving the way for Internet 3.0, the metaverse and a digital world. 

As van Rijmenam explains, “Digital is here to stay. If you don’t digitally transform your business now, you will be out of business.” 

With [Subspace GlobalTURN](https://subspace.com/product/globalturn), you can have edge-to-edge acceleration for any service you deploy. You can move beyond regional limitations and deliver your real-time traffic in real-time. 

Digital transformation is necessary to survive in today’s world. But it doesn’t have to be painful or costly. [Subspace](https://www.youtube.com/watch?v=2GZV-wZvKSo) offers a simple API, so you provision, point your traffic and go. And you get always-on DDoS protection without sacrificing speed. 

Van Rijmenam sees the potential for the Metaverse and digital transformation. “There are so many opportunities for organizations to improve communication,” he says. “It will revolutionize how we work, where we work, with whom we work.” 

The time is now. Don’t just survive the digital world. Be digital and thrive. Transform your business today when you get up and running in Subspace to make real-time work in real-time. 

Why Digital Matters Now: Dr. Mark van Rijmenam on the Metaverse

**TL;DR**

Repeated outages of popular services, including Roblox, Facebook and Amazon Web Services, is a warning to us all about the fragility of IT infrastructure as it exists today. Real-time applications need a better internet. See how Subspace can improve your app by [trying Subspace GlobalTURN for free now](https://console.subspace.com).

Estimated read time: 7 minutes

------------------------------

During the past year, more people around the world have been online than ever before. There have also been major outages, each drawing attention to the dependence on the aging internet.

In April, [Microsoft suffered an outage](https://techcrunch.com/2021/04/01/microsoft-outage-knocks-sites-and-services-offline/) that took Xbox and its Office cloud services offline. Six months later, [Facebook and its related apps](https://techcrunch.com/2021/10/04/facebook-messenger-instagram-whatsapp-are-all-down/)—Facebook, Messenger, Instagram and WhatsApp—all went dark. Both of those outages were related to the failure of a Domain Name Server (DNS), the core naming structure that forms the infrastructure for the web.  Later that same month, while most of the tech world focused on Facebook’s announcement of its [rebranding as Meta](https://subspace.com/resources/more-than-metaverse-subspace), the popular online gaming platform Roblox, which hosts such games as Adopt Me! and Jailbreak, announced a [“service disruption”](https://roblox.fandom.com/wiki/2021_Roblox_outage) affecting 99% of its users. This disruption continued for 73 hours. In early December 2021, [Amazon Web Services then went down](https://techcrunch.com/2021/12/07/amazon-web-services-went-down-and-took-a-bunch-of-the-internet-with-it/), bringing down not only shopping on Amazon, but also crippling Associated Press, Disney+ and Vice along with hundreds of smaller companies that use AWS for backend services. 

*Users say Amazon Web Services is suffering a major outage. The company provides cloud computing services to individuals, universities, governments and companies including The Associated Press.
The Associated Press on Twitter (@AP)*

# Everyone is Susceptible to IT Outages

These outages are not unique. Every company has (or should have) a response plan ready to be deployed in the event of an IT outage. They know that the plan is not there for “if” the outage occurs, but “when” the outage occurs.

*[Roblox’s] days-long predicament serves as a reminder that beneath all the hype, the metaverse as it exists today is built on problem-laden infrastructure.
[Mathew Olson](https://www.theinformation.com/reporters/mathew-olson), technology journalist and author of the Reality Check Newsletter*

**30-Year-Old Technology is Not the Answer**

The increasing outages we have seen in 2021 should serve asRoblox is the canary in the coal mine, warning all of us that building the future of global communications on top of 30-year-old technology is not a workable solution. The world needs something better, built for how we use online applications today.

![](https://www.datocms-assets.com/41207/1639676426-image3.png)

The [internet](https://subspace.com/resources/subspace-is-an-internet-built-for-real-time) was not built for today’s real-time applications. The US Department of Defense conceived the internet to connect contractors and research stations scattered around the country, enabling them to share information. The priority for the network was reliability over speed. The result was the mesh of networks stitched together by routers using protocols with built-in retry logic. This network could survive natural and manufactured disasters—but it was slow.

A significant upgrade to the internet came in the 2000s with the introduction of the World Wide Web. The focus at that time was delivering masses of static data such as rich websites, high-resolution images, and streaming video. The push resulted in a massive increase in bandwidth delivered to homes, but the underlying impediments remain.

# Modern Real-Time Applications Need a Better Network

The latest wave of web applications is interactive and real-time. Real-time applications such as video conferencing, remote desktop, and VoIP require a network with no latency. Even when the internet is working as designed, it may not perform well enough for real-time applications. A delay of two or three seconds is acceptable when a user is waiting for a web page to load, but that same delay feels like an eternity when your video chat has frozen. Because of the intrinsic internet design, a packet of data may travel thousands of miles out of its way and pass through many routers, firewalls, and other devices before reaching its destination. Each hop on the packet’s journey introduces additional latency, jitter, and packet loss. The milliseconds of delay accumulate and contribute to a poor user experience.

These challenges occur when the internet is operating as designed. When you add in the certainty of failed hardware, a power outage, or a fiber cut, you go from performance problems to a complete outage. For many companies, an outage can cost millions of dollars per hour and, for all companies, there is a loss of user confidence. For example, during the Roblox outage, the company’s  competitor, [Rec Room](https://twitter.com/ShawnRecRoom/status/1454535046090530818), saw massive downloads for the service’s apps and record highs for concurrent players.

**An IT Outage Can Ruin Careers**

An outage like this can ruin careers for IT managers, especially if the outage could have been avoided. This situation is even scarier when you consider that internet failures are out of the hands of most corporate IT managers. The solution is to move traffic to a new network designed for real-time applications. That network is [Subspace](https://subspace.com).

# Subspace is the Better Network

In addition to resolving most of the public internet’s performance issues, Subspace can prevent or mitigate many network outages. Subspace uses a proprietary, AI-driven [weather map of the internet](https://subspace.com/resources/why-internet-weather-isn-t-going-anywhere-and-what-it-means-for-remote-work). Internet conditions are updated multiple times per second. Subspace uses this data to find the fastest path for an application’s data. It is like Waze for the internet. An application on Subspace enjoys minimal latency, jitter, and packet loss in normal times and routes around disruptions such as malfunctioning routers and cut lines just like Waze would route you around an accident or closed road.

![](https://www.datocms-assets.com/41207/1637001413-howsubspaceworks.png)

# DDoS Attacks Also Cause Outages

Subspace also protects servers from DDoS attacks, another common cause of service outages. Subspace has DDoS built into the network. It scans for malicious packets in real-time and removes them from the network. As the Subspace AI determines the path of the attack, scrubbing moves to the node closest to the attack source and blocks the attack from the network. The result is that application servers don’t need to deal with a deluge of spurious messages, and firewalls don’t get bogged down.

![](https://www.datocms-assets.com/41207/1639681445-screen-shot-2021-12-16-at-2-03-28-pm.png)

For IT managers responsible for creating or operating real-time applications, they just can’t afford to trust their business to 30-year-old technology. Subspace is a requirement, not an option. 

Contact [Subspace](https://www.subspace.com/contact) now, before the next outage strikes.

Lessons on Outage Prevention

**TL;DR**

In-game voice chat played a key role in the development of multiplayer gaming. Voice deepens the immersive world, helps forge social bonds, and strengthens online play. Now, as gaming is quickly blending into the ascending metaverse, real-time communications matter more than ever, which means both gaming and metaverses need a network capable of enabling them. See how Subspace can improve your voice app by [trying Subspace WebRTC-CDN for free now](https://console.subspace.com/try).

Estimated read time: 10 minutes

------------------------------

From its earliest implementations, voice chat has been critical in driving multiplayer gaming evolution. Before widespread broadband and Voice over Internet Protocol (VoIP) availability, players  needed to game over their data connection and communicate over a separate phone line, which would be tied up for the duration of game play. Things began to change in 1995, with the integration of VoIP technology into MechWarrior 2: 31st Century Combat, which had a networked player vs. player mode under Windows 95. Sega had numerous Dreamcast games with voice chat in 1999, and then Microsoft introudced voice into their Xbox. Voice would become pivotal in the success of Microsoft’s subscription-based Xbox Live service.

On the PC side, third-party applications, such as Ventrilo, TeamSpeak, and Xfire began to offer voice services that ran alongside multiplayer games. Some of these applications gained impressive functionality over time. For example, by the end of the early 2000s, Xfire had added support for Google Talk, Facebook Chat, and even in-game video broadcasting. Many PC game developers didn’t bother to integrate VoIP into their titles because players were already using these third-party tools (although there were plenty of exceptions, beginning with Dungeons and Dragons Online in 2005). And upon its public release in 2015, Discord blew every other gaming-centric chat app out of the water, in no small part because it was high-quality, scalable to 5,000 simultaneous users, and free. Microsoft provided Discord integration for Xbox Live in 2018. While Discord’s focus has broadened beyond gaming, it remains the favored solution for gaming voice chat, as evidenced by Sony bringing [Discord to the PlayStation Network](https://www.gamedeveloper.com/console/sony-invests-in-discord-plans-to-integrate-discord-into-psn-next-year/#open-in-new-tab) in 2022.

![](https://www.datocms-assets.com/41207/1639497618-image3.jpg)

A decade ago, University of Melbourne student Gregory Robert Wadley submitted his 313-page PhD thesis titled “[Voice in virtual worlds](http://hdl.handle.net/11343/37224).” Despite its age, the project remains arguably the best summation of how real-time voice technology influences, elevates, and occasionally inhibits amazing online entertainment experiences. It’s a sweeping analysis with extensive context and insight on how modern use of voice can enhance multiplayer gaming and the coming metaverse.
Yet many people resisted voice when it first began to replace text in online multiplayer games, in part because users felt self-conscious about their voices and how their voice might be at odds with the persona they wanted to project.

Modern voice modulation technologies can change speech to fit the user’s preferences in real time. Wadley found that “voice transforms the user experience of virtual worlds” and “makes some forms of collaboration more efficient.” Spatial voice, in which speakers in a virtual space are perceived in terms of direction and proximity, “increases immersion, reduces channel clutter, and affords new strategies for team coordination.” In short, when deployed well, voice can radically improve the social elements of virtual existence…provided the software and underlying infrastructure can deliver the performance necessary to fulfill voice’s promise.

# Why in-game voice chat matters to the gaming experience

But having voice chat in games isn’t always desirable. Technically, it can be difficult to achieve satisfactory real-time voice performance across multi-platform games, although [Fortnite’s Party Hub](https://www.theverge.com/2019/9/18/20871464/fortnite-voice-chat-party-hub-houseparty/#open-in-new-tab) is one exemplary exception. Bringing in voice and background sounds from the real world can break the fantasy illusion of a gaming universe (especially parental hollering). Some players find the auditory chaos of public game lobbies confusing and overpowering (and potentially abusive). In situations where players on a team have not developed group habits or discipline, players may talk over each other and inhibit satisfying gameplay.

In most cases, though, players find in-game voice chat to be both more enjoyable and additive to in-game effectiveness, as when working as a squad in tactical first-person shooters. Voice alleviates the need to type messages in text chat, allowing both hands to stay on the keyboard/mouse and better keep players in the immediate game experience. Socially, voice in games can heighten fun, suspense, and even fear. It also enables easier conversation, allowing strangers to become acquaintances, and allows friendships to form that may even extend offline.

Voice also brings humanity into the gaming experience. Sometimes, that humanity distinguishes other in-game entities from AI-based non-player characters and makes the world more immersive. Voice makes the fun of in-game trash talk more visceral and engaging. (Try trash-talking your Alexa device for an exercise in futility.) Better engagement translates to more time playing. Ali Nhari, head of sales for audio/video SDK firm Agora, [notes](https://www.linkedin.com/pulse/why-in-game-videovoice-chat-important-every-game-genre-ali-nhari/#open-in-new-tab), “It’s not only about strategy, shooter, or role-playing games. Casual and card games, for instance, are directly benefiting from In-game Video/Voice chat, increasing their daily active users by at least 20% with double session time!”

# Voice chat is transitioning multiplayer games to the metaverse

Everyone with two working eyes knows the benefits of stereoscopic vision — seeing the world in 3D depth rather than the strangely flatter world witnessed with only one eye. Similarly, we perceive stereo sound as richer and more lifelike than mono. The benefits continue with added channels and speaker locations. So-called surround sound via 5.1, 7.1, Dolby Atmos, and DTS:X grow increasingly close to reality, albeit with more hardware, complexity, and cost.

Since the 1990s, companies have sought to mimic surround sound through as few as two speakers (as with headphones) by using complex auditory algorithms. Put simply, it’s possible to model how sounds from various sources penetrate the ear canals. Imagine someone drops a ball several feet away to your right. The sound made by the bounce reaches your right ear before your left. It also reflects off walls and other surfaces, so it will reach your ears differently. These varying paths can be simulated in a model known as head-related transfer function (HRTF). The best results require calibration to the specific user, and headphones can yield more consistent results than desktop speakers due to how the user changes position within the speakers’ sound field. When done well, 3D spatial audio can sound like it’s coming from any direction (including above, below, and behind the user) and any distance.

![](https://www.datocms-assets.com/41207/1639497613-image2.jpg)

In a field that benefits from immersion, the leap from stereo to spatial audio feels like the jump from mono to stereo. However, the computation needed to create and play spatial audio is formidable, and today, that means requiring a dedicated audio chip to handle that computation so the CPU remains free for other tasks. Fortunately, gaming hardware manufacturers are increasingly ready for this. Oculus has supported 3D audio in its API since the Rift headset. The PlayStation 5 and Xbox Series X|S have hardware-enabled 3D audio for their PS5 3D, and Windows Sonic for Headphones products, respectively.

The implications of spatial audio for gaming and in-game voice chat should be crystal clear. In conventional shooter games, the sound of gunfire from a rooftop sniper sounds like it’s coming from the ground. Spatial audio fixes this. In horror games with low visibility, when the user relies more on audio cues, spatial audio can be more informative and deliciously more ominous. Similarly, when a teammate shouts, “Get down!” the user has a better sense of the other player’s direction and distance. In fact, in any game environment, from FPS to casino blackjack, spatial audio will provide a greater sense of environmental immersion, improved communication clarity, and closer bonding to nearby characters. It also allows game developers to limit perception of other voices to only nearby speakers “within hearing range,” thus cutting down on that previously mentioned environmental confusion.

As you might guess from Oculus’s early integration of 3D audio, it’s only a small hop from the virtual worlds of conventional gaming to a bona fide metaverse. Battle royale titan PUBG incorporates spatial audio, which helps with team mechanics, and now PUBG creator Krafton is [talking about](https://afkgaming.com/esports/news/krafton-ceo-reveals-plans-to-create-a-pubg-metaverse/#open-in-new-tab) expanding the game into a “complete interactive virtual world.” In September 2021, Roblox followed suit, [announcing](https://blog.roblox.com/2021/09/future-communication-metaverse/#open-in-new-tab) an in-game voice chat feature along with a new spatial voice beta for developers, all wrapped up in the context of Roblox being a metaverse in which “people can come together within millions of 3D experiences to learn, work, play, create, and socialize.” In Roblox’s mind, these elements are inextricably linked. The object is to let users communicate in the most natural, convenient, and immersive ways possible.

# Impact of in-game voice chat on the metaverse

In [Part III](https://www.matthewball.vc/all/networkingmetaverse/#open-in-new-tab) of his “The Metaverse Primer,” Matthew Ball discusses differences between a “mirrorworld” like Microsoft Flight Simulator and a more conventional game like Roblox. Flight Simulator is based on 2.5 petabytes of information (potentially even including real-time) pulled straight from the real world. The titanic size of this virtual world prohibits keeping more than a sliver of it on a local system. During play, mass amounts of data must stream in from remote servers, filling in the world the user is about to enter just before he or she reaches it. On the other hand, Roblox keeps most world data on the client device and primarily updates incidental data, such as object positions. The burden falls on local storage and computation rather than the network.

![](https://www.datocms-assets.com/41207/1639497628-image6.jpg)

The existence of these two metaverse models raises several questions. With Roblox, changes within the world are temporary and only seen by a relative few people. This involves far less world-level resource load than having all changes be permanent for all users forever. Which model is better for a metaverse? If the answer is some hybrid mix, how will that be handled? And what happens if and when metaverses overlap, yet they use different permanence models?

As so often happens in “this or that” technology questions, the answer is “both.” We’ll need local compute and storage for impermanent worlds, and exceptional network performance for permanent and/or mirrorworld ones. And keep in mind that online games based in the former model already strain for performance across conventional network links. Latency — the delay between transmission and reception between two network entities — currently bedevils many games and real-time communication applications, because even a split-second delay will ruin the experience. Closer to the user, network bandwidth becomes a potential issue, as the resolution and refresh rates of next-generation virtual worlds can dwarf the bandwidth needs of, say, high-def streaming video. And not least of all, reliability must persist from the user to the application’s core servers and all network exchange points in between. In a world of congestion, cut lines, and misconfigured equipment, this is easier said than done.

The point is that the network bottlenecks that have challenged real-time communications for decades will not only persist but magnify as gaming evolves into even more resource-intensive metaverses. In-game voice chat is an essential, enabling component of bridging games into metaverses, but it is only one of many such components. Metaverses of any stripe will need sufficient network performance to enable real-time communications that span the real and virtual worlds concurrently. Very little distinction will soon exist between games and virtual worlds in this regard.

This rising need is why Subspace is now rolling out, as [VentureBeat](https://venturebeat.com/2021/10/26/subspace-will-launch-its-parallel-and-real-time-internet-for-gaming-and-the-metaverse/#open-in-new-tab) calls it, the company’s “parallel and real-time internet service for gaming and the metaverse.” This will be a network-as-a-service offering for developers, especially for those making real-time games set in massive virtual worlds. Subspace is a direct response to the public internet’s technical shortcomings and their impacts on gaming and communication experiences. The company’s globally deployed network operates alongside the public internet and addresses those needs around latency, bandwidth, and reliability.

![](https://www.datocms-assets.com/41207/1639497622-image5.png)

Tomorrow’s [metaverse](https://subspace.com/resources/metaverse) will likely be made by game publishers. Some of those publishers are already giving better gaming experiences to millions of new users on Subspace’s network. As the benefits of those improved experiences drive more revenue, Subspace will continue to expand its infrastructure, so the metaverse will have a network foundation able to meet any emerging model or need.

What Are the Trends Shaping the Future of Multiplayer Gaming?

Why In-Game Voice Chat Became Essential to Multiplayer Gaming and How It Impacts the Metaverse

Giacomo Vacca

Duke Hong

David Swafford

Nikki Shum-Harden

Daryll Swager

Nathan Moore

Mo Nezarati

William King

Matt Ringel

Subspace Team

Contact Centers

WebRTC-CDN

Network Measurement

FAQs

Documentation

Round-Up

Voice

WebRTC

Remote Work

Metaverse

SIPTeleport

RTPSpeed

Dev Spotlight

PacketAccelerator

Network Control

Sports Betting

Video Games

Communications

products

solutions

resources

developers

pricing

company

Case studies

Quickstart Guides

White Papers 

Blog

Videos

Subspace is here for WebRTC applications.

navigation_mobile_cta

The world’s fastest internet for real-time applications—period. Every millisecond counts. Learn more in our newsletter.

Subspace is here for WebRTC applications.

WebRTC-CDN

SIPTeleport

PacketAccelerator

GlobalLoadBalancer

Network Optimization

Blog

Resources

Documentation

About us

Subspace is here for WebRTC applications.

Blog

Subspace Solves Challenges of Real-time Pathfinding...

Subscribe to our newsletter

Subspace Appoints Davide Petramala Vice President of Sales...

Why Today’s Internet Can’t Support Remote Contact Centers...

SpaceX is Giving the Internet Lift With Starlink...

ICE Negotiation...

How to Measure Network Performance...

STUN 101: Establishing WebRTC Sessions Around Firewalls...

VoIP Jitter: Guide for Testing, Measuring and Reduction...

How Gather is Revolutionizing Video Conferencing...

Why Digital Matters Now: Dr. Mark van Rijmenam on the Metaverse...

Lessons on Outage Prevention...

Why In-Game Voice Chat Became Essential to Multiplayer Gaming and How It Impacts the Metaverse...