Capacity Enhancement Of Cellular Network Using Wifi Offloading

The growing popularity of bandwidth-intensive applications as well as the proliferation of mobile devices in Malaysia has resulted in the accelerated growth of mobile data traffic or also known as the mobile data explosion phenomena. Hence, cellular network providers are struggling to keep pace in providing the demand of network capacity. Since the deployment of WiFi hotspot apparently is more cost-effective than upgrading the network infrastructure, WiFi offloading approach is an alternative that offers the most feasible solution to support cellular networks. This thesis developed a quantitative study on the performance of mobile data offloading through WiFi networks. 100 Android users from various backgrounds were recruited, and a real-time statistics of user’s WiFi utilization were collected for 18 days of measurement study using a customized application development. WiFi offloading performance parameters such as temporal and spatial coverage has been studied in depth by conducting a meticulous demographic analysis using few developed algorithms. Findings reported only about 16% WiFi temporal coverage across all users, which reveals that users approximately utilize an average of 4 hours WiFi daily, and indicates higher utilization for student compared to a public group. The acquired whole-day traces also indicates WiFi coverage distribution through spatial coverage analysis which shows dense coverage distribution at residential area, especially between 1400 to 2300 hour. On the other hand, the study also takes into consideration the user’s associated WiFi network performance parameters, that reveals low packet loss and moderately skewed throughput analysis. In addition, an analytical study on a single server using Markovian model has been deployed to observe the impact of WiFi offloading on cellular network capacity. In the case of without WiFi offloading deployment, the queue behaviour demonstrates higher traffic intensities when customer arrival rate is high, and therefore decrease cellular network capacity, and vice versa. To the best of our knowledge, this is the first measurement study that carried out a real data trace of user’s daily WiFi offloading in the context of Malaysian cellular network subscriber. It shows that WiFi offloading approach has a positive prospect in augmenting the cellular network as well as providing new insights for network planning, policy or creative price plans