To solve the problem that traditional Wireless Body Area Network (WBAN) prediction model has low prediction accuracy, large computational complexity and high energy consumption, an adaptive cubic exponential smoothing algorithm based on penalty error matrix was proposed. Firstly, a lightweight prediction model was established between the sensing node and the routing node. Secondly, blanket search was used to optimize the parameters of the prediction model. Finally, penalty error matrix was used to further refine the parameters of the prediction model. The experimental results showed that compared with the ZigBee protocol, the proposed method saved about 12% energy in 1000 time slot range; compared with blanket search method, the prediction accuracy was improved by 3.306% by using penalty error matrix. The proposed algorithm can effectively reduce the computational complexity and further reduce the energy consumption of WBAN.

To resolve the problems currently existing in access protocol for Terahertz Wireless Personal Area Network (THz-WPAN) for PicoNet Coordinator (PNC) handover process, such as imperfect PNC handover process, redundant information transmission, waste of obvious timeslot resources, an access protocol of piconet coordinator high efficient handover for terahertz wireless personal area networks called PCHEH-AP (Piconet Coordinator High Efficient Handover Access Protocol) was proposed. By using deleting the specified node redundancy information mechanism, confirming first and handover next mechanism, PNC handover adaptive slot allocation mechanism, PCHEH-AP generally regulates the PNC handover process, improves channel utilization, reduces data delay, and makes PNC handover more reasonable and efficient. The simulation results show that compared with the protocol defined by IEEE 802.15.3c and the HTLL-MAC protocol, the data delay of PCHEH-AP is reduced by at least 8.98% and the throughput of MAC layer is also increased by 3.90%.