|
As we embark on a new year and decade, it seemed worthwhile to take a peek at the principal forums for global 5G industry technical collaboration and do a quick assessment of what is occurring and who are the “leaders.” The leadership dimension is especially relevant in Washington these days—which is suffering from a peculiar 5G dementia. As the year ended, there were no less than 35 current 5G related Congressional legislative actions, several of which actually passed one of the chambers. One is literally named “promoting United States international leadership in 5G.”
These efforts in an alternative reality are orchestrated by a recently appointed loyal White House drone with numerous adhoc K-Street lobbying organizations smoking the 5G dope produced around town. (Cannabis is now legal in Washington.)
5G Leadership Reality
There are many ways to measure leadership in the 5G arena, but one of the most compelling and measurable is to examine the participation in the principal industry technical venues where all the parties collaborate together in deciding 5G features and instantiating them in global specifications that everyone implements in products and services worldwide. Those parties consist of chipset vendors, equipment manufacturers, service providers, network operators, support organizations, institutes, and government agencies.
Participation is also a considerable undertaking because it requires the necessary resources to understand the subject matter and what is occurring across the dozens of bodies that are meeting now almost every month. The principal defining body for 5G—3GPP—now has 1,114 work items (435 legacy, 601 full 5G, 78 next-generation 5G). The full “stand-alone” 5G specs are known as Release 16 and scheduled for adoption this year. The participation metrics of both attendance and substantive submissions are the ultimate reality test of who are the serious parties in the real 5G world. Less measurable are the “bonus points” for leadership acquired by initiating and leading particularly innovative new work and 5G features.
The meetings for advancing this work are hosted among the participants, and now typically occur every month—constantly rotating around the world at locations in Asia, Europe, and the Americas. In addition to advancing the work, the meetings also serve as a critical means for obtaining consensus among the participating parties, for establishing features and implementation schedules, for discovering and serving customers, and demonstrating substantive leadership. The processes are well-honed and have been highly successful globally in developing the world’s most significant network communications medium—the global mobile communications infrastructure.
January was an unusually slow 5G collaborative “breather” month—presaging a considerable array of meetings in February. It was marked by meetings of the 5G architecture group (SA2) in Incheon, the CODEC group (SA4) in Wroclaw, the mission-critical communications group (SA6) in Hyderabad, and the core E2E network & terminals group (CT1) virtually. So, who was present? Who contributed input documents to shape the results—arguably the most important metric of leadership?
The 5G Architecture group is one of its most important and active. It’s interim meeting in Incheon attracted 194 participants from 82 industry and government players, with the largest number from Huawei, Samsung, ETRI, Ericsson, LG, Nokia, and InterDigital. A number of other U.S. players were present, including both DHS’ FirstNet and the DOJ. There were 1831 document submitters from 72 industry players, led by Huawei/HiSilicon, Nokia/Nokia Shanghai-Bell, Ericsson, Qualcomm, Samsung, and Vivo. Among government agencies, DHS FirstNet had 3. The principal work item submissions included 38 key 5G architecture features, at the top of which were Network Automation for 5G, Proximity-based Services, Vertical and LAN Services, advanced V2X, Edge Computing, Non-Public Networks, Cellular IoT support and evolution for the 5G System, Access Traffic Steering, Switch and Splitting support, Policy and charging control, 5G multicast-broadcast services, multi-USIM devices, Specific services support, Location Services, Enablers for Network Automation for 5G, Enhancement of Network Slicing, and Enhancements to the Service-Based 5G System Architecture.
The 5G CODEC group is a highly specialized group that attracts those who offer multimedia platforms. Its Wroclaw meeting attracted 67 participants from 32 industry players, with the largest numbers from Ericsson, Qualcomm, Nokia, Huawei, InterDigital. Notably, Dolby, Apple, and even Facebook were present. There were 166 document submitters from 23 industry players, led by Qualcomm, Ericsson, Sony, Dolby, and Nokia.
The 5G mission-critical group is also highly specialized that deals with both National Security Emergency Preparedness (NSEP) capabilities as well as high resilience capabilities such as required for utility infrastructures. Its Hyderabad meeting attracted 68 participants from 33 industry and government players with the largest numbers from the Indian government, Samsung, ZTE, AT&T, Huawei, Motorola, and Qualcomm. There were 196 document submitters from 30 industry and government players, led by Samsung, Huawei/Hisilicon, Ericsson, ZTE, Convidia, and Nokia/Nokia Shanghai Bell. The Netherlands Police was the only government agency submitting contributions. The principal work item submissions included 13 key 5G mission-critical features, at the top of which were enabling Edge Applications, Mobile Communication System for Railways, application layer support for Factories of the Future, Mission Critical Services support over 5G System, architecture and information flows for Mission Critical Data, application layer support for V2X services, support for Unmanned Aerial System (UAS), Mission Critical services over 5G multicast-broadcast system, and location enhancements for mission-critical services.
The 5G E2E core network and terminals group met virtually among 35 participants from 23 industry and government players, including DHS FirstNet, with the largest numbers from Nokia, Qualcomm, Ericsson and MediaTek. There were 172 document submitters from 14 industry players with the largest numbers from Huawei/HiSilicon, Ericsson, Nokia/Nokia Shanghai Bell, MediaTek, Qualcomm and Blackberry. Almost all contributions were focussed on a single work item—the NAS protocol, which is used to manage the establishment of communication sessions and for maintaining continuous communications with the user equipment as it moves.
5G Security
In the ever-important world of 5G security, some of the most significant work continues to be led by the UK government. The assertion out of Washington that somehow China controls or dominates the 5G security activity flies in the face of myriad security activities that proceed on a collaborative consensus basis, and at which the principal U.S. government security agencies and regulators have long been absent, much less active contributors.
In addition, some of the details of the annual Security Week event emerged with a focus on 5G security and an announcement of a 5G security certification initiative.
Sponsored byVerisign
Sponsored byRadix
Sponsored byVerisign
Sponsored byWhoisXML API
Sponsored byCSC
Sponsored byDNIB.com
Sponsored byIPv4.Global
One of the confusing things about the phrase “5G” is that it is often lumped with millimeter wavelength technology.
With my technical hat on I am very concerned about the ability of network stack implementations to handle the increased side effects that will occur due to the use of ever higher frequencies, shorter ranges, and more frequent base-station/tower transitions.
Those side effects are likely to manifest themselves as bursts of lost packets; increased packet duplication and re-ordering; and definitely as variation in packet latency (i.e. jitter.)
Those things are often not handled well by protocol stacks. We can anticipate a lot of user applications, especially those based on UDP, to show problems - or even fail - as a result. (TCP implementations may also have difficulties.)
There is an enormous amount of misinformation about this.
My company builds gear that acts as a man-in-the-middle device to introduce the kinds of packet issues that I mentioned. Our customers use that to make sure that their code is robust as network conditions get sub-optimal. (At the risk of being a bit commercial, our company is at IWL And many of our customers have been asking "How is 5G going to affect us?" Our answer is somewhat vague - that the two new bands in 5G are going to affect how radio penetrates and reflects and will probably end up meaning that you will be transitioning bdetween base station/towers more frequently. None of us yet have a really solid sense of how to characterize those transitions except to know that there will almost likely be changes in delay before and after, and that during the transition there may well be a lot of jitter and packet delivery issues (including duplication and re-ordering.) That problem gets a lot worse as we move to millimetre bands. Those are not strictly part of the formal 5G definition from 3GPP, but it seems that various carriers are sliding it in without clearly saying so. In those higher frequency (shorter wavelength) bands we can expect even more effects - like significant changes in delay or jitter, or packet loss, simply if a person using a mobile phone turns around, walks into another room, or another person walks by. (Remember how old TV was affected by airplanes flying at low altitudes overhead?) Too many people who write code don't understand, or worse, don't test, on networks that are undergoing these kinds of things. We've kind of gotten used to audio dropouts and video artifacts, but imagine them much worse. And with the advent of more and more things in vehicles on roads (Vehicle-to-X interactions) and on higher speed trains, the issues of transitions will likely get worse. I've seen some vehicle-to-vehicle stuff that used 5G (without intervening cell towers) so that vehicle A can report to vehicle B what it is seeing about vehicle C (which may be hidden from vehicle B). That's a typical car approaching on a blind slide-road situation. It's going to be hard to isolate and test all of the problems in code that could happen due to combinations of packet delivery issues such as delay and especially duplication or re-ordering. When I was doing similar things with video, even seemingly trivial differences, like a millisecond or so variation in delay in arrival of two different packets, could cause code to fall over and become a quivering lump.