Cross-layer modeling for quality of service guarantees over wireless links

Abstract

Abstract-We propose a cross-layer approach to investigate the impact of physical-layer infrastructure on data-link-layer quality-of-service (QoS) performance over wireless links in mobile networks. At the physical layer, we take multiple-inputmultiple-output (MIMO) diversity schemes as well as adaptivemodulation-and-coding (AMC) techniques into account. At the data-link layer, our focus is on how this physical-layer infrastructure influences the real-time multimedia delay-bound QoS performance. To achieve this goal, we first model the physical-layer service process as a finite-state Markov chain (FSMC). Based on this FSMC model, we then characterize the QoS performance at data-link-layer using the effective capacity approach, which turns out to be critically important for the statistical QoS guarantees over wireless links in mobile networks. We also investigate the impact of physical-layer power control and channel-state information (CSI) feedback delay on the QoS performance. The numerical results obtained demonstrate that our proposed cross-layer model can efficiently characterize the interactions between the physical-layer infrastructure and datalink-layer QoS performance.Index Terms-Cross-layer design and optimization, mobile wireless networks, quality-of-service (QoS), effective capacity, adaptive-modulation-and-coding (AMC), multiple-inputmultiple-output (MIMO), real-time multimedia delay-bound.

Cross-Layer-Model Based Adaptive Resource Allocation for Statistical QoS Guarantees in Mobile Wireless Networks

“…Thus, we assume uncoded transmission for analytical convenience. Second, based on our study in [16], we observe that the performance trends of FEC/ARQ-based transmission is similar to that of uncoded systems, as long as the link BER is not too high. Therefore, the investigation of the uncoded system also provides a guideline on designing the coded system.…”

“…Based on our application of the effective capacity method [16], [17], the system resources are allocated according to the heterogeneous fading channel statistics, the diverse QoS requirements, and different traffic characteristics. Specifically, our scheme adaptively assigns power-level and time-slots for real-time mobile users in a dynamic timedivision multiple access (TDMA) mode to guarantee the bounded delays.…”

“…Furthermore, when delay-bound is the main QoS metric of interest (i.e., when the focus is on delay-bound violation probability), an expression similar to Eq. (15) can be obtained as (16) where D max denotes the delay-bound, and δ is jointly determined by both arrival and service processes, which will be elaborated on below.…”

Channel-aware MAC performance of AMC-ARQ-based wireless systems

“…For instance, the target link layer PLR P lmax can only be guaranteed with a small violation probability , that is, Pr{P u l ≤ P lmax } ≈ κ u e −ψ * u P l max ≤ , where ψ * u is the unique real solution of [26]. It is worth stating at this point that except for low input data rates, the tail probability tends to overestimate the packet loss probability and it can only be used as long as A S .…”

Interference Mitigation in CR-Enabled Heterogeneous Networks

“…On the other hand, the effective capacity, denoted by E C (θ), is a duality of the effective bandwidth, which specifies the maximum constant arrival rate that can be supported by the system subject to a given θ [84]. If θ * can be found as the solution of E B (θ * ) = E C (θ * ), δ can be obtained by [87] as…”