iCellular: Define Your Own Cellular Network Access on Commodity Smartphones

Leveraging multi-carrier access offers a promising approach to boosting access quality in mobile networks. However, our experiments show that the potential benefits are hard to fu lfill due to fundamental limitations in the network-controlled design. To overcome these limitations, we propose iCellular, which allows users to define and intelligently select their own cellu lar network access from mult iple carriers. iCellular reuses the existing device-side mechanis ms and the standard cellular network procedure, but leverages the end device’s intelligence to be proactive and adaptive in multi-carrier selection. It performs adaptive monitoring to ensure responsive selection and min imal service disruption, and enhances carrier selection with online learn ing and runtime decision fault prevention. It is deployable on commodity phones without any infrastructure/hardware change. We implement iCellular on commodity Nexus 6 phones and leverage Project-Fi’s efforts to test multi-carrier access among two top US carriers: T-Mobile and Sprint. Our experiments confirm that iCellular helps users with up to 3.74x throughput improvement (7x suspension and 1.9x latency reduction etc.) over the state-of-art selection. Moreover, iCellular locates the best-quality carrier in most cases, with negligib le overhead on CPU, memory and energy consumption. networks at the end device. In practice, most regions are covered by more than one carrier (say, Verizon, TMobile, Sprint, etc. in the US). With mult i-carrier access, the device may intelligently select among carrier networks and improve its access quality. To this end, industrial efforts have recently emerged to provide 3G/4G multi-carrier access via universal SIM card, including Google Project Fi [29], Apple SIM [16], and Samsung e-SIM [27]. The ongoing 5G standardization also seeks to integrate heterogenous network technologies [39]. However, our study shows that the full benefits of multiple carrier access can be limited by today’s cellular design. We examine Google Project-Fi with two carriers (T-Mobile and Sprint), and discover three problems, all of which are independent of implementations (§3): (P1) the anticipated switch never occurs even when the serving carrier’s coverage is really weak; (P2) the switch takes rather long time (tens of seconds or minutes) without service availability; and (P3) the device fails to choose the high-quality network (e.g., selecting 3G with weaker coverage rather than 4G with stronger coverage). It turns out that, these problems are rooted in the conflicts between legacy 3G/4G roaming design and user’s multi-carrier access requests. With the single-carrier scenario in mind, the 3G/4G design places the controllability of carrier access to the network side. Roaming to other carriers is not preferred unless the home carrier is unavailable. As a result, today’s carrier selection mechanism (i.e., PLMN selection) passively monitors other carriers after losing home carrier service, and selects the carrier based on pre-defined roaming preference given by the serving carrier network [13, 14]. Although viable in the single carrier case, this design limits user’s ability to explore multiple carriers. The user could miss the highquality carrier network, delay the switch with redundant carrier scanning, and get stuck in the low-quality carrier. While this problem may be solved in future architecture design (e.g. 5G), it takes years to accompolish. Instead, we seek to devise a solution that works in today’s 3G/4G network, in line with ongoing industrial efforts,