Cross-layer design of multi-rate wireless networks based on link layer truncated ARQ

A Markov chain-based queueing model is developed for a cross-layer performance analysis of a wireless network based on adaptive modulation and coding (AMC) at the physical layer and truncated automatic repeat request (ARQ) error control at the data link layer. This model generalizes previous results by using a first-order two-dimensional Markov chain to model the wireless fading channel. Furthermore, it allows each frame at the physical layer to contain an arbitrary number of packets from the data link layer and still ensuring that, concerning the acknowledged packets, the transmit buffer operates in first-in first-out (FIFO) mode. Analytical expressions for performance metrics such as throughput, average packet delay and packet loss rate, both due to buffer overflow and due to exceeding the maximum number of allowed retransmissions, are derived. These expressions are then used to formulate a constrained optimization problem to maximize the system throughput under the prescribed QoS constraints.