As a network that communicates human health information, iWNSNs (in-vivo Wireless Nanosensor Networks) must be able to detect and report back to human health in real time, and low delay and high reliability are required in iWNSNs. In view of the unstable link state in iWNSNs, a multi-hop routing protocol based on link state prediction called MRLSP is proposed. The protocol proposes a link state estimation based on Kalman filtering. In addition, an ultrasonic energy harvesting system is introduced to intermittently charge nanonodes. In the routing process, the protocol first ensures the forwarding direction of the data packet by limiting the range of the next hop node, and uses the historical state and current information of the link to estimate the link state. Finally, a fuzzy system is established using the link state, the distance and the node’s remaining energy. We use the fuzzy logic output as the criterion to select the next hop node to ensure that fewer hops complete the forwarding and successful transmission of the packet. The simulation results verify that MRLSP has advantages in end-to-end delay and data packet transmission success rate, and can maximize network performance while ensuring sufficient node energy.

作为传达人类健康信息的网络，iWNSN（体内无线纳米传感器网络）必须能够实时检测并向人类健康报告，并且iWNSN要求低延​​迟和高可靠性。针对iWNSN中链路状态不稳定的问题，提出了一种基于链路状态预测的多跳路由协议，称为MRLSP。该协议提出了一种基于卡尔曼滤波的链路状态估计。另外，引入了超声能量收集系统以间歇地给纳米节点充电。在路由过程中，协议首先通过限制下一跳节点的范围来确保数据包的转发方向，并使用链路的历史状态和当前信息来估计链路状态。最后，使用链接状态建立一个模糊系统，距离和节点的剩余能量。我们使用模糊逻辑输出作为选择下一个跃点节点的准则，以确保更少的跃点完成数据包的转发和成功传输。仿真结果验证了MRLSP在端到端时延和数据包传输成功率方面均具有优势，并可以在确保足够的节点能量的同时最大化网络性能。