Wireless Sensor Network Architecture for Monitoring and Control in Buildings

PI: Myung Lee, the City College of New York


Executive Abstract

In the U.S., buildings consume 70% of all electricity, up to 50% of which is wasted. Thus, the one of the key thrusts of planning sustainable buildings is the resource conservation while being environmentally sensible. Many components of sustainable buildings are tightly networked to monitor and control to achieve the optimal conditions for occupants of the buildings and energy expenditure.  Thousands of sensors can monitor everything from motion and temperature to humidity, precipitation, occupancy and light. The monitored information will be used to control HVAC, light, electric motors, water valves, etc. We propose to perform researches on networking all these monitoring and control devices. Our approach will be based on the wireless sensor network (WSN) with the capability to meet the various application requirements such as time delay, reliability, and scalability. When wireless devices are densely populated in the license-free frequency bands (e.g., 2.4GHz), the wireless interferences will limit the performance of WSN. Specifically, we propose to develop a novel mesh routing algorithm to provide the data delivery in a reliable and timely manner. Unlike the current practice of using a common single channel, all 16 available channels in IEEE 802.15.4 will be used to mitigate wireless interferences as well as to meet a given latency requirement. The challenge is to establish end-to-end routing paths with a sequence of optimal time slots and channels. With the allocated time slot and channel, each device in the network will be able to wake-up only when needed, substantially extending the device‚Äôs and network lifetimes. It is a difficult problem to establish end-to-end paths with latency guarantees in the presence of multiple time slots and channels. This is in general an NP-complete problem. We will investigate meta-heuristics approach to cope with challenging computational complexity.  In particular, we will consider Genetic algorithm (GA), Simulated Annealing (SA), and Particle Swarm Optimization (PSO) techniques and their variants. We will demonstrate the mesh routing protocol for a simple smart building scenario with light sensors and switches.

Objective 1: Develop an optimal multi-time and multi-channel scheduling algorithm using metaheuristic optimization.
Objective 2: Develop a novel mesh routing algorithm with multi-time slots and multi-channels
Objective 3: Develop a demo testbed for a simple smart building application.