應用於大量物聯裝置的個別傳輸與群體傳輸協定

Abstract

物聯網的出現，帶來為數眾多的工業化與個人化應用，例如: 工業監控、健康管理、智慧家庭…等。物聯網裝置常以規則性的週期來收集或回報資料給基地台。這類型的資料通常具有以下特性: (1) 大量連線裝置 (2) 小資料傳輸 (3) 固定或可預測的傳輸週期。然而，目前的行動網路溝通機制並不完全適合這類型的資料傳輸。舉例來說，透過隨機存取的競爭程序來傳送這些小數據資料會非常沒有經濟效益。在本文中，我們提出 一套有效率的小資料傳輸協定方式，以符合這類型的資料傳輸需求。我們分別定義六個不同的傳輸情境，如下: (1) 個別資料上行傳輸 (2) 個別資料下行傳輸 (3) 相同資料的小規模群體傳輸(4) 相異資料的小規模群體傳輸 (5) 相同資料的大規模群體傳輸和 (6)相異資料的大規模群體傳輸。我們所提出的解決方案基於一個創新的雜湊函式方法來進行有效率的資源分配。物聯網裝置根據自己的識別碼進行雜湊後的結果來得知被分配到的無線資源。因此，裝置便不需要進行傳統的隨機存取過程來競爭無線資源，從而節省網路頻寬及降低延遲。除此之外，我們提出的方法省去傳輸相較於小數據來說較大的識別碼也能提升整體網路的資源使用率。最後，模擬結果驗證我們所提出的協定方法對於小資料傳輸具有高效率的表現。

The emergence of Internet of Things (IoT) has enabled many industrial and consumer applications. IoT devices usually collect and report data to base stations at regular patterns. The data transmission between IoT devices and a base station normally has the following characteristics: (1) massively connected, (2) small payload per packet, and (3) periodical or at least mostly predictable. Current communication protocols, however, are not well designed to handle these characteristics. For example, contending to send a small packet by typical random access is not cost-effective. In this work, we propose a set of efficient small data transmission protocols. Six scenarios are considered: individual uplink/downlink transmission, small number of groups transmission with same/different data, large number of groups transmission with same/different data. Our solutions utilize a novel hash function to carry resource allocation information to devices. To save signaling overhead, the hash function only takes small broadcast data as input. We show that this can save a lot of contention among IoT devices are compared to traditional random access approaches. In addition, our solutions eliminate the transmission of device IDs, thus increasing throughput significantly when payloads are small. Through extensive simulations, we validate the efficiency of our solutions.