| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
--- This message came from the IEEE 802.11 Working Group Reflector ---G’day 802.11 WG,
The IEEE 802.11 Coexistence SC has been monitoring issues related to Wi-Fi coexistence, especially in the 5 GHz and 6 GHz bands, for the last 5 years. Its work has been very successful in so far as it has had an important influence (directly or indirectly) in avoiding the worst coexistence outcomes for Wi-Fi:
- LTE-U, which had some undesirable coexistence properties, appears to no longer have any market relevance, even in the US
- The MulteFire Alliance is a now mainly a marketing organisation for 5G rather than an organisation specifying technology with questionable coexistence with Wi-Fi
- Both 3GPP specified LAA and NR-U use EDCA-like access with Wi-Fi compatible timing parameters, albeit with unsophisticated “listening” based on energy detection only, rather than the more diverse Wi-Fi approach combing energy and preamble detection at the PHY level with NAV based reservations at the MAC level
- ETSI BRAN Harmonised Standards for both 5 GHz (EN 301 893) and 6 GHz (EN 303 687) are based on access requirements very much aligned with the “Wi-Fi way”
However, the coexistence situation is still not resolved and there are a variety of ongoing risks, particularly related to coexistence between Wi-Fi and LAA/NR-U in the 5/6 GHz bands. While handwaving and simulations suggested coexistence would be reasonable on “average”, it turns out that many environments are not “average”. The University of Chicago’s measurements of real (not “average”) LAA deployments suggest that LAA sometimes ignores Wi-Fi, at least in the environments that were measured. It appears the underlying issue is that the use of energy detection as the sole mechanism by LAA for coexisting with Wi-Fi does not work very well, probably due to classic hidden station issues.
Of course, these measurements only matter if LAA is widely deployed. If LAA is not widely deployed, we can all breath a big sigh of relief and get on with making the next generations of Wi-Fi even better than today. The evidence so far suggests that LAA might not be as widely deployed as some hoped. The number of operators who have deployed it in the last three years has not changed from 9 (globally), and it is not clear they have deployed it widely. The number of operators considering LAA has also remained steady for the last three years at 27 (globally). Given the slow deployment of LAA, it is possible that the 5G community is waiting for NR-U. It is quite likely that NR-U will have similar coexistence issues as LAA, but again this will only matter if NR-U is widely deployed.
At the Coex SC meeting in July 2022, this situation was highlighted, along with some questions. There were few answers in July, but answers to these questions will help direct the future work of the Coex SC. This e-mail is requesting members to think about these questions, and, at the very least, be ready to discuss potential answers at the next Coex SC meeting in Sept 2022. An even better outcome would be submissions from multiple members in relation to these questions.
The questions:
- Do measurements suggest a significant coexistence issue?
- Is there consensus that the Uni of Chicago’s measurements are a reasonable refection of coexistence between LAA and Wi-Fi?
- If not, is anyone planning to make alternative measurements for coexistence between LAA and Wi-Fi?
- Is anyone planning to make measurements for coexistence between NR-U and Wi-Fi?
- Is any coexistence issue likely to be common in practice?
- Is wide LAA deployment expected in some or all markets?
- Is wide NR-U deployment expected in some or all markets?
- Can likely coexistence issues be easily mitigated?
- Are there any obvious mitigations if coexistence issues, like those shown by the Uni of Chicago measurements, do arise in practice?
- Is avoidance (ie Wi-Fi and LAA/NR-U always operate on different channels) a workable coexistence mitigation? Is this approach feasible in 500 MHz (the current situation in Europe) or does it require 1200 MHz (the current situation in the US)
I would like to run a session at the meeting in September 2022 where members discuss these and similar questions (hopefully based on additional submissions). I then plan to run a series of straw polls to better understand the perspectives of the membership and so set a platform for future progress.
Please contact me if you plan to make a submission or if you have any thoughts on these questions …
Andrew Myles
To unsubscribe from the STDS-802-11 list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11&A=1
To unsubscribe from the STDS-802-11 list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11&A=1