Search

Participant selection method is one of the important contents of crowd sensing research. Existing research still has some shortcomings. Only the attributes such as task time or task area coverage are considered, which makes the selected participants perform tasks less efficiently. Therefore, in order to comprehensively consider the task time and task area coverage constraints, this paper proposed a selection method of crowd sensing participant based on the greedy ant colony algorithm ( PSGACO) to achieve the highest task execution efficiency and the minimum incentive cost of the crowd sensing platform. The method mainly selected the participants who were suitable for performing the publishing task by the concentration of the ant pheromone concentration of the candidate participants, and greatly improved the task execution efficiency. Finally, it compared the proposed PS-GA CO method with the common participant selection method through simulation experiments. The experimental results show that PS-GA CO is superior to the other two methods in terms of algorithm running time, task execution efficiency and incentive cost, and has a good application prospect for the crowd sensing participant selection. [ABSTRACT FROM AUTHOR]

Mobile crowd sensing (MCS) is the core of mobile computing with the rapid development of wireless sensor networks and the popularity of mobile intelligent devices. The use of crowd sensing can be completed large-scale, complex environment and social awareness tasks, which the task distribution is an important part of this application. In order to solve the problem that the perceived environment is complex, the number of users can not meet the requirements and the quality of the collected data is low, this paper proposed a task distribution mechanism EUC based on social attribute and effective user calculation. According to the task to filter the characteristics of the user, from the user point of view, EUC considered the user’s social, the user’s social network through the relevant information to increase the number of effective users. From the platform point of view, according to the task of receiving and submitting the situation, EUC dynamically adjusted the effective user’s points,thus ensuring the effective number of users of the entire system. Theoretical analysis and experimental results show that the proposed mechanism can improves the task distribution efficiency of the system and improves the quality of the collected data. [ABSTRACT FROM AUTHOR]

The crowd sensing realizes the maximization perception for crowd, which has been considered the key techno-logy for society network. However, to perceive the position of the participants and trajectoiy uncertainty problem caused by poor group of mental perception data real-time, this paper proposed a crowd sensing group of intellectual task allocation algorithm based on space task allocation. The algorithm adopted the dynamic and adaptive data-driven scheme to obtain the optimal model to solve the problem of dynamic perception. Track algorithm was based on public mobile model (based on Markov model), Bayesian inference, according to the initial task to estimate the next position, data acquisition strategy based on this algorithm could achieve a local service to guide the future data collection, thus it completed the entire loop feedback of perception. Task assignment is proved based on the uncertain path of mobile group of mental perception task allocation is effective. [ABSTRACT FROM AUTHOR]

For the past few years, the continued in-depth study of crowdsourcing calculation, social computing and opportunistic network in wireless sensor network area, there put forward a new sensing model called crowd sensing while its theory and key technology have an opportunity for further breakthrough. Crowd sensing constitutes a network of people-centered, social behavior analysis become an important data collection method under this human centric network. This paper explored the origin of crowd sensing, delved into the development further and analysed its research significance ultimately. It summarized three aspects on the present research status closely related to platforms, social attributes and opportunistic data collection through adopting the data collection technology form social behavior analysis, then analyzed the relationship between the three, came up with critical issues that future work might facing. Finally it made a conclusion to hopefully play a guiding role on the direction of follow-up study. [ABSTRACT FROM AUTHOR]

In the existing community detecting methods of the crowd sensing, the weight and direction of the edges were often not considered. This paper proposed a community detection algorithm based on node optimal path (CDOP) to realize the reasonable division in weighted and directed network by calculating the optimal path tree (OPT), the similarity index (SI) and community discrete index (DS). Experimental results show that for the MIT data, the accuracy of algorithm is up to 94. 2%, which is higher than the 10. 9% in other models. In addition, the algorithm has good accuracy in different types of network data sets as well, so further illustrates the effectiveness of the algorithm. [ABSTRACT FROM AUTHOR]