EFFICIENT APPROACH OF CHANNEL STATE INFORMATION PREDICTION FOR 5G NETWORKS USING CROWD SENSING OPTIMIZATION ALGORITHM
Publication Date : 31/12/2018
In mobile crowdsensing (MCS), among the participants’ main concerns is the cost for 5G data usage, which influences their willingness to take part in a Crowdsensing task. In this paper, present the look and implementation of an MCS info uploading mechanism CSI prediction in reducing additional 5G data cost incurred by the complete crowd of sensing participants. By taking into consideration the two most common real-life 5G price plans-unlimited data plan (Unlimited) and pay as you go (Payout PER), CSI prediction partitions all of the users into two teams corresponding to these two price plans at the beginning of each month, with the aim of minimizing the total refunding budget for all individuals. The partitioning is founded on predicting users’ mobility patterns and sensed info size. The CSI prediction mechanism was created influenced by the observation that during the data uploading cycles, Unlimited users could opportunistically relay Shell out PER users’ info to the crowdsensing server without extra 5G price, pro- vided both types of users will be able to “meet” on a common native cost-no cost network (e.g., Bluetooth or WiFi direct). To carry out the experiments using both Massachusetts Institute of Technology fact mining and the tiny World In Action (SWIM) simulation data sets. Analysis results exhibit that CSI prediction could lessen total 5G data expense by up to ∼50%, in comparison with the direct-assignment approach that assigns each participant to Unlimited or PAY PER directly according to the size of her sensed information.
No. of Downloads :