Abstract
With the popularization of wireless technology, more and more wireless portable devices can be found in our life, such as mobile phone, laptop, PDA, etc. People in the present day can send or receive messages and get information from network at any time and any where with these wireless portable devices. But all these wireless portable devices have the same problem, the limited power lifetime. Since the mobility characteristics, most wireless portable devices are powered by lifetime-limited battery. Therefore, how to save power plays an important role on the design of a wireless communication system. Power saving can be achieved by hardware and software. By hardware, for example, the design of a power efficient hardware component can save the power consumption during the operating time. By software, an enhancement of the wireless communication protocol with power-awareness is also desirable. For example, the discontinuous reception (DRX) strategy, which is adopted in many wireless communication systems, is a popular power saving strategy. This strategy periodically wakes up a sleep mobile device to listen to incoming messages. However, the previous DRX strategy suffers from drawbacks since all mobile devices, which may process different characteristics, must have the same sleep interval. This would result in inefficient power management. Therefore, in the paper, we introduce an improved DRX strategy that overcomes the limitations of original DRX strategy and saves more power. By the proposed DRX strategy, mobile devices can select an individual sleep interval freely and thus make the power management more flexibly and efficiently. Furthermore, network needs not to keep track of each mobile device’s sleep interval, but schedules all downlink data with only one sleep cycle. Besides, the side effects caused by the proposed strategy are bound and determinable. For example, the proposed strategy can save half of the power consumption on receiving and processing paging signals with an increase of two retransmission on average for one request and 1.5 seconds paging delay.