Minimum Length Scheduling With Packet Traffic Demands in Wireless <italic>Ad Hoc</italic> Networks

Abstract

Traditional approach to the minimum length scheduling problem ignores the packet level details of transmission protocols, meaning that a packet transmission can be divided into several data chunks each of which is transmitted at a different rate due to the difference in the set of concurrently transmitting nodes. This solution requires including packet headers for each data chunk resulting in both an increase in the system overhead and underutilization of the time slots. In this paper, we extend the previous works on minimum length scheduling by considering the transmission of the packets of arbitrary sizes in the time slots of arbitrary lengths. Given the packet traffic demands on the links, we formulate the joint optimization of the power control, rate adaptation and scheduling for minimizing the schedule length of a wireless ad hoc network and demonstrate the hardness of this problem. Upon solving the power control and rate adaptation problem separately, we formulate the scheduling problem as an integer programming (IP) problem where the number of variables is exponential in the number of the links. In order to solve this large-scale IP problem fast and efficiently, we propose Branch and Price Method and Column Generation Method based heuristic algorithms.