Maximizing Sleeping Capacity Based on QoS Provision for Information-Centric Internet of Things

Unlike the IP-address-centric network, heterogeneous devices in the information-centric Internet of Things (IC-IoT) have heterogeneous quality of service (QoS) requirements. Employing sleeping technology in IC-IoT while satisfying the heterogeneous QoS requirements for IoT devices is challenging. In this paper, considering the IoT devices' heterogeneous QoS requirements, we investigate a sleeping capacity maximizing problem in IC-IoT. First, a content-based heterogeneous network architecture is proposed to empower IC-IoT and decouple the heterogeneous content transmission functions. Specifically, macro base stations (MBSs) are used to guarantee the control coverage and transmit the low QoS requirement services, whereas small base stations (SBSs) are utilized to transmit the high QoS requirement services. The QoS requirement levels are differentiated by the outage constraints which are related to the achievable data rate. Then, an optimization problem is formulated to maximize the sleeping probability and available bandwidth while the outage probability and achievable data rate constraints are guaranteed. In the formulated optimization problem, sleeping probability plays the roles of both an optimization variable and an optimization objective, which makes the optimization problem challenging to be solved. In order to decouple sleeping probability from the optimization values, we first analyze the parameters related to the sleeping probability and the available bandwidth and then propose a bisection-based nonlinear programming algorithm. This algorithm divides the original problem into two subproblems. One is related to the bandwidth reallocation. The other one involves obtaining the optimal sleeping probability. Last but not least, analytical results are concluded due to the simulation results, which indicate the following facts: 1) optimal sleeping probability is linear to the available bandwidth, and inversely proportional to the devices' density, and; 2) sleeping probability is more sensitive to the changes of channel quality than those of QoS provision.