Project Goals

Internet-of-Things (IoT) or Machine-to-Machine (M2M) communication is fast becoming a significant characteristic of a wide spectrum of networked applications. With ‘smart’ sensing devices becoming ubiquitous, the need to support large-scale IoT communication will be a reality in the near future and IoT traffic will also need to be supported over cellular networks.

IoT devices and smartphones share the same network infrastructure, but current cellular data networks are primarily designed, engineered, and managed for smartphone usage, and the typical interactive applications used by people. Moreover, the protocol framework to support IP traffic has evolved from the cellular network’s initial circuit-switched/admission-controlled protocol framework developed to support voice traffic.

Tunnels are established between the user device (often termed UE for user equipment) and the cellular network’s network data center for each session, which is a significant overhead and is unsustainable for IoT communication. It is important that the cellular network not maintain large amounts of state on a per-device basis. This challenge can only be dealt with by a careful examination and optimization of tradeoffs between different end-system and in- network functions associated with IoT communication. The project seeks to investigate novel mechanisms to eliminate cellular network tunnels (between mobile devices and the network gateway).

The project seeks to also design a new transport protocol that can dynamically adapt based on IoT application requirements, and is flexible enough to effectively handle large data transfers, short transfers supporting a send-and-forget paradigm, and ensuring that these can efficiently coexist with traditional smartphone traffic, all while supporting mobility.