Trust Verification Method for Leveraging Secure Task Completion in Mobile Crowd Outsourcing Processes

Mobile crowdsourcing (MCS) is an autonomous paradigm that integrates humans, technology, devices, and computing methods for cost-effective task completion. The tasks are outsourced from service providers, organizations, etc. for ease of completion. Such a process is concerned with trust and privacy-dependent security considerations. This article provides a solution for mitigating illegitimate workers causing malicious injections in heterogeneous task processing instances. The method named Rendezvous Trust-enabled Task Processing (RTTPM) recommends workers and devices based on individual highness value. This highness represents the trust observed from the previous task completion intervals. In this trust computing and verification process two distinct phases of Q-learning are utilized; the transitions between false data injection and task completion are independently validated for highness value derivations. Learning offers a best-fit reward with updates based on the maximum possible highness derivatives from the transitions. If the derivatives are diverging from the rendezvous maximum reward, then the worker/device is malicious, and its trust is nullified. Such nullified workers/devices are thwarted from the current task handling interval preventing further malicious injections. The reward computation is halted for the same from which a new transition phase is initiated. This process is active until all the tasks are processed and their precise completion is maximized.