基于差分總變化量的時(shí)變圖信號(hào)重構(gòu)算法

江瑞

摘? 要： 圖信號(hào)具有一定的網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)，通過(guò)網(wǎng)絡(luò)上的節(jié)點(diǎn)收集相關(guān)信號(hào)的數(shù)據(jù)。但是在數(shù)據(jù)收集過(guò)程中由于圖信號(hào)具有時(shí)變特性，容易造成數(shù)據(jù)損失或錯(cuò)誤。文中主要研究無(wú)權(quán)無(wú)向的時(shí)變圖信號(hào)網(wǎng)絡(luò)，對(duì)時(shí)變信號(hào)數(shù)據(jù)進(jìn)行差分處理可以使得信號(hào)表現(xiàn)出更好的平滑特性。提出采用差分總變化量的方法對(duì)時(shí)變圖信號(hào)數(shù)據(jù)重構(gòu)，并使用均方根誤差方法對(duì)重構(gòu)結(jié)果做評(píng)價(jià)，通過(guò)實(shí)驗(yàn)仿真發(fā)現(xiàn)基于差分總變化量的方法可以取得較好的重構(gòu)效果，從而驗(yàn)證了采用差分總變化量的可靠性。

關(guān)鍵詞： 時(shí)變圖信號(hào); 差分總變化量; 信號(hào)重構(gòu); 數(shù)據(jù)恢復(fù); 數(shù)據(jù)差分處理; 重構(gòu)結(jié)果評(píng)價(jià)

中圖分類(lèi)號(hào)： TN911.7?34? ? ? ? ? ? ? ? ? ? ? ? ?文獻(xiàn)標(biāo)識(shí)碼： A? ? ? ? ? ? ? ? ? ? ? ? 文章編號(hào)： 1004?373X（2020）13?0053?04

Time varying graph signal reconstruction algorithm based on

difference total variation

JIANG Rui

（MOE Key Laboratory of Cognitive Radio and Information Processing， Guilin University of Electronic Technology， Guilin 541004， China）

Abstract： Graph signals possess a certain network topology， and its collection of the relevant signal data is by network nodes. However， it is prone to result in data loss or error due to the time?varying characteristics of graph signals in the process of data collection. In this paper， the unweighted and undirected time?varying graph signals network is studied mainly. The difference processing of time?varying signal data can make the signal show its better smoothing characteristics. In this paper， the method of total difference variation is proposed to reconstruct the time?varying graph signal data， and the root?mean?square error method is used to evaluate the reconstruction results. It is found in the experimental simulation that the method based on the total difference variation can achieve good reconstruction results， which verifies the reliability of using the total difference variation.

Keywords： time?varying graph signal; difference total variation; signal reconstruction; data recovery; data difference processing; reconstruction result evaluation