One of the key promises of 5G and Beyond 5G wireless networks is to provide connectivity for a massive number of Internet of Things (IoT) devices, in order to support emerging applications such as e-health, smart home, and industrial internet. However, such massive connectivity poses an enormous challenge on how to coordinate uplink data transmissions with a limited amount of spectral resources. First, since each device generates intermittent data with very short message payload, the traditional data communication protocols, e.g., 4G-LTE, are not suitable for IoT traffic due to expensive signaling overhead before data transmission. Second, many such applications are time-senstive, and thus not only require high throughput, but also low latency and high information freshness. Therefore, it remains an open challenge to develop low-overhead, highly spectrum-efficient, and low-latency communication protocols.
Our recent work has tackled this challenge in two fronts. First, we have developed new wireless control algorithms that can exploit the MIMO performance gain for uplink massive access, without colleting CSI or collecting queue length information. As a result, these algorithms can significantly reduce the overhead for massive IoT access, at a level independent of the number of interfereing devices in the system. Second, we study how to schedule data sources in a wireless time-sensitive information system with multiple heterogeneous and unreliable channels to minimize the total expected Age-of-Information (AoI). Although one could formulate this problem as a discrete-time Markov Decision Process (MDP), such an approach suffers from the curse of dimensionality and lack of insights. Instead, we developed a new partial-index policy, which signficiantly generalizes the classic Whittle's index, and leads to decomposed decisions with complexity growing only polynomially with the number of devices.
Selected publications: