Application of Genetic Algorithm in Time-Based Wireless Sensor Network Schedule Optimization

In recent years, wireless sensor networks (WSN) have been particularly interested, studied and applied very strongly. A sensor network is generally limited in resources and energy, which greatly restrict its applicability. Sensor network optimization in practice is a very diverse with a wide range of applications, whereas sensor network scheduling is important in lowering energy consumption and maximizing network lifetime. However, optimization of sensor network schedule a very complex problem with many constraints that is not trivial to solve by analytical methods. This article discusses a heuristical approach using a genetic algorithm to find an optimal solution for network scheduling. The evaluation of fitness function, as well as selection with crossover and mutation operations help to evolve individuals in the population through generations in an optimal direction.

An Incentive-Compatible Combinatorial Auction Design for Charging Network Scheduling of Battery Electric Vehicles

Charging network scheduling for battery electric vehicles is a challenging research issue on deciding where and when to activate users’ charging under the constraints imposed by their time availability and energy demands, as well as the limited available capacities provided by the charging stations. Moreover, users’ strategic behaviors and untruthful revelation on their real preferences on charging schedules pose additional challenges to efficiently coordinate their charging in a market setting, where users are reasonably modelled as self-interested agents who strive to maximize their own utilities rather than the system-wide efficiency. To tackle these challenges, we propose an incentive-compatible combinatorial auction for charging network scheduling in a decentralized environment. In such a structured framework, users can bid for their preferred destination and charging time at different stations, and the scheduling specific problem solving structure is also embedded into the winner determination model to coordinate the charging at multiple stations. The objective is to maximize the social welfare across all users which is represented by their total values of scheduled finishing time. The Vickrey–Clarke–Groves payment rule is adopted to incentivize users to truthfully disclose their true preferences as a weakly dominant strategy. Moreover, the proposed auction is proved to be individually rational and weakly budget balanced through an extensive game-theoretical analysis. We also present a case study to demonstrate its applicability to real-world charging reservation scenarios using the charging network data from Manhattan, New York City.

A Study of BLE Mesh Network Scheduling Algorithm Using A Smartplug

This paper proposes a BLE Mesh network scheduling algorithm using SmartPlug. And proposes random-backoff, a very simple method for collision avoidance. The low-energy scheduling algorithm of SmartPlug and BLE node is proposed. SmartPlug periodically broadcasts its own information to inform the surrounding BLE nodes and SmartPlug. A node for data communication with SmartPlug is decided for efficiency. The BLE node periodically transmits data and switches to sleep mode after data transmission. In this paper, propose a very simple method, random-backoff, to reduce collisions when transmitting data in a BLE node. In the simulation of results shows when the data size is 1 ~ 31 bytes and 32 ~ 255 bytes, and random-backoff is effective when 1~31 bytes length. In the case of 32 to 255 bytes, random-backoff is not effective, so additional research is needed. And the simulation conducts under the same condition, but it shows that the data length and the number of transmission attempts has an effect. The simulation results show the transmission success ratio is similar regardless of the data length when random-backoff is not applied. It also shows the number of transmission attempts has an effect. Also, this simulation shows the results that as the number of nodes increases, the wireless environment becomes congested, and the transmission success ratio decreases. As a result of the simulation, random-backoff for collision avoidance is effective in transmitting data of 1 to 31 bytes better than 32 to 255 bytes in the data length.