Mobile crowd computing (MCC) has emerged as an ideal platform for accessing required computing services from public-owned mobile devices in the vicinity, not requiring to going to the Cloud. But it may happen that the service is not available within the network (of the consumer) always. In this case, a carrier is needed, which carries the request to the service provider (in another network), gets the service from it and handovers to the consumer. But the mobility of the service consumer, provider, and the carrier poses a great challenge in binding between them for service request and service exchange. Device mobility is considered as one of the performance metrics in MCC. Though it is not trivial to measure, the efficacy of MCC heavily depends on this metric. To mitigate this issue, in this paper, we propose a service provisioning model in MCC based on the mobility patterns of the above-mentioned three entities. We applied a mobility prediction algorithm on the UCSD dataset that comprises real-traces of 235 mobile device users for 78 days across 402 access points (APs). In this experiment, we focussed mainly to get the information: (a) average time gap after a user connects to an AP, (b) average duration he/she remains connected to an AP, and (c) a set of users who remains connected to a particular AP simultaneously. Knowing the mobility patterns of the service consumer, provider, and the carrier, in terms of the above-mentioned information, is helpful to bind them in particular time frames. This allows avoiding the liveness problem (consumer waits for the service indefinitely) and availability problem (carrier returns with the service but cannot find the consumer).