++++++
In this new edition, we cover the main technical elements that help define the value of latency in a telecommunications network. The field is vast, and the subject remains the cornerstone of much research into how to minimize it, or at least control and predict it, in order to make the user experience the best and therefore the smoothest possible. There is no miracle recipe, but a succession of segments that contribute to its value.
At the IBC2024 (International Broadcast Convention) in Amsterdam, the subject of low latency video/streaming was very much on the agenda. For example, Haivision demonstrated its SST protocol compared with open source SRT, and the Comcast US/Ateme FR/Castlabs US consortium presented the results of their work on ultra-low streaming (LL-DASH). See also https://www.svta.org/ Streaming Video Association.
Introduction
Latency is a key parameter to monitor for real time connectivity for applications, either for B2B enterprise (M2M, remote control, …) or B2C consumers segment (gaming, video…).
In a non-trivial network, a typical packet will be forwarded over multiple links and gateways, each of which will not begin to forward the packet until it has been completely received. In such a network, the minimal latency is the sum of the transmission delay of each link, plus the forwarding latency of each gateway. In practice, minimal latency also includes queuing and processing delays.
Latency’s Round Trip Delay ranges from below ms to hundred ms
High Speed Trading: 0.1 ms (Server to Server)
Industry Ethernet/IP: ~1 ms (M2M)
Wireless Network (Wifi, 4G/5G): 20-100ms (RAN <> Transport <> Core)
LEO satellite: < 300ms (Air <> Teleport <> CPE)
GEO Satellite: 500-800ms (Air <> Teleport <> CPE)
What about human reactions? for an easy task, the average response time would be around 200-300ms.
Figure 1: wide range of latency values
There are multiple ways to consider streamlining latency and its variation the jitter:
Distance from UE to Server for End-to-End,
The passive infrastructure used to carry the signal,
The signal itself and modulation,
Wave penetration and modulation,
The active infrastructure with the protocol deployed,
The design and topology of the network,
Protocol implemented.
Let’s review some parameters that shape latency values.
Distance & Support
Physical support is essential in the contribution to latency value and the signal propagation for wireless transmission: Fiber contribute to a 0.1 ms per km latency, copper transmit 0.5 ms per sec (transmission at 200 000 km/sec), electromagnetic waves are transmitted at 300 000 km per sec.
As a telecom network is made up of several sub-networks, each built on a specific medium, wired or wireless, the resulting end-to-end latency will depend on each segment. It will then be necessary to evaluate the weight of each segment, and the possibility of modifying or correcting the segment in question. Subsequently some bottlenecks will be identified.
Active Infrastructure
Nodes such as routers, switches, firewalling, server, CPU Load impact on the latency. Also type of ASICS, IP routing tables, security drive also latency.
Protocols at stake
The choice of protocol depends on the specific application and network constraints. Here are a few examples of commonly used low-latency protocols:
Transport: UDP (User Datagram Protocol), is suitable for purposes where error checking and correction are either not necessary or are performed in the application. UDP avoids the overhead of such processing in the protocol stack. Time-sensitive applications often use UDP because dropping packets is preferable to waiting for packets delayed due to retransmission, which may not be an option in a real-time system.
SCTP (Stream Control Transmission Protocol): network protocol that is connection-oriented and used for transmitting multiple streams of data simultaneously between any two endpoints.
SST: In computer networking, Safe Stream Transport (SST) is a transport protocol that provides an ordered, reliable byte stream abstraction similar to TCPs, but enhances and optimizes stream management to permit applications to use streams in a much more fine-grained fashion than is feasible with TCP. SST is patented by Haivision company.
SRT: equivalent somehow to SST is an open-source protocol that provides connection and control, reliable transmission like TCP; however, it does so at the application layer, using UDP protocol as an underlying transport layer. It supports packet recovery while maintaining low latency (default: 120 ms). SRT also supports encryption using AES.
DSCP: Differentiated Services Code Point (DSCP) is a means of classifying and managing network traffic and of providing quality of service (QoS) in modern Layer 3 IP networks. It uses the 6-bit Differentiated Services (DS) field in the IP header for the purpose of packet classification.
Routing OSPF (Open Shortest Path Protocol) & BGP (Border Control Protocol),
The BBR protocol runs purely on the sender and does not require changes to the protocol, receiver, or network, making it incrementally deployable. It depends only on RTT and packet-delivery acknowledgment, so can be implemented for most Internet transport protocols. See detailed.
The new L4S standard seeks to reduce the amount of congestion that network packets spend queuing in a network. See the white paper from Orange on L4S.
LL-DASH: Dynamic Adaptive Streaming over HTTP (DASH), also known as MPEG-DASH, is an adaptive bitrate streaming technique that enables high quality streaming of media content over the Internet delivered from conventional HTTP web servers.
CPU
Latency oriented processor architecture is the microarchitecture of a microprocessor designed to serve a serial computing thread with a low latency.
Conclusion
As we have just seen, the latency quality of service indicator depends on a large number of parameters. Controlling, measuring and predicting it requires a good understanding of network topology, as well as the ability to modify certain parameters (it's not impossible that Pareto's law applies), the ultimate objective being the search for a better user experience and its economic impact, as mentioned in the previous article.
Any subjects you're interested in, let's engage.
Marc Soulacroup
Tel 00 33 6 2550 7046
Paris, France
Comments