Energy Saving Algorithms for Mobile TV Using Time Slicing

Abstract

With the rapid development of mobile technologies, mobile TV has emerged and is becoming increasingly popular. However, the viewing time of mobile TV on a mobile device can be limited due to its battery-power limitation. In order to prolong the viewing time the mobile device’s energy consumption needs to be reduced. A technique referred to as time slicing has been proposed in the mobile TV standards such as DVB-H and ATSC M/H to reduce the energy consumption in mobile devices. The idea of time slicing is to broadcast video data in bursts. Consequently, mobile devices can only receive video data during the burst time and turn off their radio frequency (RF) circuits during inter burst intervals to save energy. The burst scheduling problem has some constraints. For example, in order to prevent playback interruption in the mobile device, intervals between two consecutive bursts cannot be too long. When a mobile device receives a burst, the burst data will be stored in its buffer for consumption during the next inter-burst interval. The size of the burst cannot be too large or it may exceed the buffer capacity of the receiving mobile device. Furthermore, the burst transmission increases the channel switching delay because a mobile device has to wait for the first burst to start playback. The channel switching delay is a critical issue in mobile TV services and it should be limited to a small value while designing the burst scheduling algorithm. The optimal burst scheduling algorithm for multiple arbitrary bit rate TV channels with the constraints mentioned above is NP-complete [7]. Therefore, sub-optimal burst scheduling algorithms such as Double Buffering Scheduling (DBS) and Adaptive Data Transmission (ADT) have been proposed [9]. Although these algorithms can construct a valid burst schedule, the channel switching delay is not considered and there is still a room to improve the energy saving in these algorithms. In this dissertation two burst scheduling algorithms for time slicing in mobile TV networks are proposed. The first one is called Regulated Period Scheduling (RPS) which can achieve on average about 10% higher energy saving ratio than that of the existing DBS algorithm with very low computation complexity. Since the optimal energy saving ratio is on average about 11% higher than that of the existing DBS, the energy saving ratio of the proposed algorithm in this dissertation is very close to the optimal ratio. The goal of the proposed second algorithm is to maximize the energy saving of mobile devices with a channel switching delay limit. The simulation results demonstrate that the energy saving of the proposed algorithm is just about 2% less than the optimal algorithm. This shows the performance of the second proposed algorithm is very close to that of the optimal algorithm.