Adapting to Dynamic LEO-B5G Systems: Meta-Critic Learning Based Efficient Resource Scheduling

Abstract

Low earth orbit (LEO) satellite-assisted communications have been considered as one of the key elements in beyond 5G systems to provide wide coverage and cost-efficient data services. Such dynamic space-terrestrial topologies impose an exponential increase in the degrees of freedom in network management. In this paper, we address two practical issues for an over-loaded LEO-terrestrial system. The first challenge is how to efficiently schedule resources to serve a massive number of connected users, such that more data and users can be delivered/served. The second challenge is how to make the algorithmic solution more resilient in adapting to dynamic wireless environments. We first propose an iterative suboptimal algorithm to provide an offline benchmark. To adapt to unforeseen variations, we propose an enhanced meta-critic learning algorithm (EMCL), where a hybrid neural network for parameterization and the Wolpertinger policy for action mapping are designed in EMCL. The results demonstrate EMCL's effectiveness and fast-response capabilities in over-loaded systems and in adapting to dynamic environments compare to previous actor-critic and meta-learning methods.

Deep Reinforcement Learning Approach for HAPS User Scheduling in Massive MIMO Communications

“…where RpK, K 1 , Zq " R 1 pZ, K 1 q `RpKq is the total sumrate. Using (10), for given a, we can obtain K, and K 1 . Thus, RpK, K 1 , Zq and Rpa, Zq can be used interchangeably.…”

“…Assuming the coexistence of both TBSs and NTBSs, users should be associated with the proper access network to enhance the system data rate. Consequently, many works study user scheduling in integrated terrestrial and nonterrestrial networks (between NTBSs and terrestrial cells) to integrate NTBSs into existing terrestrial wireless networks [8], [10], [11]. In [11], the resource allocation problem in a vertical heterogeneous network (VHeNet) is studied to maximize the downlink throughput of ground users.…”

Rate-Splitting Multiple Access-Based Satellite–Vehicular Communication System: A Noncooperative Game Theoretical Approach

“…the set s(t) ≜ [s c (t), s 1 (t), ..., s K (t)] are data streams, which are linearly precoded by using the Transmit Pre- K+1) . In the downlink process, we assume that Doppler shift caused by the high mobility of LEO can be predicted and compensated [45], [50], [51]. And, the Doppler shift at the vehicle is contained in the estimation error in CSIT.…”

Computation Offloading and Resource Allocation Based on P-DQN in LEO Satellite Edge Networks

“…As stated in Section 3.1 , the channel state of all terminals in the same satellite coverage is almost identical, so the satellite uplink channel state in region k at slot t can be defined as where denotes the transmit antenna gain of terminals, is the channel fading between region k and the associated satellite at slot t , and represents the receiving antenna gain of satellites. In particular, the channel fading between terminals and satellites generally includes free space path loss, atmospheric fading, and small-scale fading (obeying the Rician distribution) [ 45 ], i.e., where c represents the light speed, is the distance between region k and its access satellite at slot t , denotes the carrier frequency, specifies the atmospheric fading, and is the Rician distributed small-scale fading. More precisely, atmospheric fading is expressed as where , H is the orbital altitude of LEO satellites, and is the attenuation through rain and clouds, separately.…”