WLAN Interface Management on Mobile Devices-thesis
The number of smartphones in use is overwhelmingly increasing every year. These devices rely on connectivity to the Internet for the majority of their applications. The ever-increasing number of deployed 802.11 wireless access points and the relatively high cost of other data services make the case for opportunistic communication using free WiFi hot-spots. However, this requires effective management of the WLAN interface, because by design the energy cost of WLAN scanning and interface idle operation is high and energy is a primary resource on mobile devices. This thesis studies the WLAN interface management problem on mobile devices. First, I consider the hypothetical scenario where future knowledge of wireless connectivity opportunities is available, and present a dynamic programming algorithm that finds the optimal schedule for the interface. In the absence of future knowledge, I propose several heuristic strategies for interface management, and use real-world user traces to evaluate and compare their performance against the optimal algorithm. Trace-based simulations show that simple static scanning with a suitable interval value is very effective for delay-tolerant, background applications. I attribute the good performance of static scanning to the power-law distribution of the length of the WiFi opportunities of mobile users, and provide guidelines for choosing the scanning interval based on the statistical properties of the traces. I improve the performance of static scanning, by 46% on average, using a local cache of previous scan results that takes advantage of the location hints provided by the set of visible GSM cell towers. The number of smartphones in use is increasing every year, and sales of smartphones will overtake laptop sales in These mobile devices are equipped with multiple . Network Interface Connectors (NIC) including a WLAN interface for 802.11, and a cellular (GSM or CDMA) interface. Among the available data communication technologies on current smartphones, 802.11 has some unique features that make it the best choice for data communication. First, it is a relatively short-range radio technology, therefore it offers higher data rates and consumes less energy per byte [1] as seen in Figure 1.1. Also, using freely available organizational or home WiFi networks is preferable to communication over expensive cellular data plans. To find WLAN connectivity opportunities a smartphone should scan the environment regularly. Once a usable access point [25] is discovered the interface can associate with it and start communication. Therefore, if a device has pending data to send or receive, the WLAN interface should be active and scanning even when it is not associated with any access points. However, because of the carrier sense nature of the 802.11 standard, WLAN interfaces consume considerable energy for scanning and even when idle
With a very strong assumption, that of knowing the future WiFi availability, the algorithm to find the optimal interface schedule is simple and has polynomial running time. In a special case, where the medium is never unavailable for short intervals, a simple greedy algorithm can find the optimal strategy. There have been proposals and attempts to predict the future connectivity of a mobile user. In particular Nicholson and Noble [24] have proposed using a Hidden Markov Model to predict future connectivity of mobile users. Although they take advantage of locationing technologies, they can only reliably predict future connectivity for a few minutes (the accuracy of the prediction dramatically drops beyond two minutes). Therefore I will not follow this path of research. Instead I will investigate heuristic strategies for regularly scanning the environment. These strategies are introduced in the next chapter and briefly analyzed