Research On Next Generation WiFi Cooperative Spatial Reuse And Resource Allocation

The increasing demand for data applications in the 5G era requires wireless access networks to achieve increasingly higher throughput and millisecond low latency.Examples include applications such as 4k/8k video,AR,VR and cloud gaming.The existing Wi Fi 6802.11 ax standard cannot meet these demands.The chief goal of the next generation Wi Fi 7802.11 be standard is to further improve Wi Fi throughput and reduce latency in dense deployment scenarios of access points(AP).It introduces some new technologies at the physical and media access control(MAC)layers,its core technologies including multi-AP coordination and multiple resource unit(MRU).For the overlapping basic service set(OBSS)scenario,it is worthwhile to study how to increase the spatial reuse opportunities of multiple basic service sets(BSS)to improve the total system throughput.The reasonable setting of carrier sense threshold(CST)value of multiple BSSs is the key to realize concurrent transmission of multi-BSS systems.However,the existing 802.11 ax standard and 802.11 be Draft0.3 do not address the specific algorithm to adjust the CST.Most of the research results focus on the adjustment of CST for single BSS.On the other hand,unlike the 802.11 ax standard which supports only one resource unit(RU)assigned to one user,802.11 be supports the technique of assigning multiple RUs to one user,but there are many constraints on the combination of specifications of multiple RUs.Therefore,it is important to study the multi-RU resource allocation algorithm under the 802.11 be standard to improve the system throughput and resource utilization.There are no related research results yet.This thesis focuses on the above two key issues,and the main work consists of two parts.First,the spatial reuse problem of multi-AP collaboration is studied for the OBSS scenario.It designs a multi-AP collaborative spatial reuse algorithm based on reinforcement learning,models the multi-AP collaborative spatial reuse problem in the presence of master APs as a Markovian decision process,and constructs the corresponding state space,action space and corresponding reward functions for different business type scenarios,respectively.Secondly,it designs the interaction steps of collaborative learning and collaborative spatial reuse between master-slave APs,and proposes a multi-AP collaborative adjustment of CST spatial reuse strategy.Simulation results show that the proposed strategy in this paper can effectively improve the throughput performance by about 47% and reduce the throughput jitter compared with the scheme using a fixed CST in traditional protocols.Then,this thesis surveys the multi-RU technology of 802.11 be and designs an algorithm for 802.11 be based multi-RU resource allocation.Given the four different service types defined in the 802.11 e standard and the restrictive conditions of 802.11 be Draft0.3 on the combination of multi-RU specifications,it defines user satisfaction based on service types to establish an objective function,defines user priority in terms of channel conditions and service types of users,and schedules users with high priority in a greedy manner first.It proposes an algorithm to match multiple RU specifications for users in conjunction with user rate requirements.In this paper,the problem of matching station(STA)and RU locations is modeled as a problem of solving complete matching in an assignment bipartite graph,and the matching strategy between STA and RU locations is based on the classical kuhn-munkras(KM)algorithm of graph theory.Simulation results show that the proposed scheme in this paper improves throughput performance by about 14% and user satisfaction by about 35%compared to the baseline scheme that divides RU resources as equally as possible.