CosMac Overview and Goals

Our goal is to identify bottlenecks in large-scale IoT-picosat constellations and design new primitives to eliminate those bottlenecks. Specifically, we aim to achieve the following:

• Scale: We aim to support a large number of satellites (100s of satellites) and IoT devices (100k devices).

• Performance: We aim to maximize the end-to-end throughput for IoT traffic.

• Robustness: We should be able to recover from transient failures/errors at satellites or ground stations.

• Low-power: We should incur little additional power overhead for IoT devices and picosats.

Fig. 2c shows an overview of CosMAC’s design. As shown, CosMAC’s uplink medium access algorithm (Sec. 4.2) runs on each IoT device. The rate of data upload, i.e., the flow rate, is mediated by our flow control algorithm (Sec. 4.3) running on picosats that communicate control information to IoT devices through existing periodic beacon transmissions. The centralized scheduler, operating on the cloud, determines the optimal schedule for downlink transmissions from picosats (Sec. 5). This schedule is then communicated to the picosats through TT&C-capable ground stations. Subsequently, the downlink transmissions adhere to this schedule and can be received at one or more ground stations. The received data from the ground stations is forwarded to the cloud.

我们的目标是识别大规模物联网-皮卫星星座中的瓶颈，并设计新的 原语（primitives） 来消除这些瓶颈。具体而言，我们的目标是实现以下几点：

• 可扩展性：我们旨在支持大规模的卫星（数百颗）和物联网设备（十万台）

• 性能：我们的目标是最大化物联网流量的端到端吞吐量

• 鲁棒性：系统应能从卫星或地面站的瞬时故障/错误中恢复

• 低功耗：我们应当只给物联网设备和皮卫星带来极少的额外功耗开销

图2c展示了CosMAC设计的概览。如图所示，CosMAC的 上行链路介质访问算法 （第4.2节）运行在每台物联网设备上。数据的上传速率（即流量速率）由运行在皮卫星上的流量控制算法（第4.3节）进行调节，该算法通过现有的周期性信标传输向物联网设备传达控制信息。运行 在云端的中心化调度器 负责确定来自皮卫星的 下行链路传输的最优调度方案 （第5节）。该调度方案随后通过具备TT&C（遥测、跟踪与控制）能力的地面站传达给皮卫星。之后，下行链路传输将遵循此调度方案进行，并可被一个或多个地面站接收。从地面站接收到的数据最终被转发到云端。