深圳大学学报理工版

随着无人机(unmanned aerial vehicle, UAV)的广泛应用,其与地面接收站实时共享机载传感器数据成为工业界的迫切需求.然而,目前为无人机开放使用的频谱资源稀缺,通信信道带宽非常有限,这就促使人们研究如何在带宽受限条件下实现无人机数据的实时无损回传.作为突破经典奈奎斯特(Nyquist)采样理论的新技术,压缩采样将是解决这类问题的最佳方案.本文通过对比当前无人机通信技术核心参数,揭示现有通信技术标准无法满足无人机数据通信对信道带宽日益迫切的需求,评述无人机数据回传压缩技术,对压缩感知、1-bit压缩采样、相位恢复和矩阵补全技术原理进行回顾.采用压缩感知和矩阵补全技术对实测数据进行验证,结果表明,压缩感知和矩阵补全技术可在带宽不变的情况下,显著降低数据的传输时间.最后提出无人机数据压缩和恢复领域的4个研究发展方向.

With the widespread application of unmanned aerial vehicle(UAV), the real-time sharing of data between UAV and base station has become an urgent demand in industry. However, the spectrum resources available for UAV data transmission are extremely precious, resulting in rather limited channel bandwidth. This, in turn, motives ones to explore efficient technologies for real-time non-destructive backhaul of drone data under bandwidth-constrained conditions. As a new technology breaking through the classical Nyquist sampling theorem, the compressive sampling(CS)turns out to be a promising solution to the aforementioned problem. By comparing the core parameters of current UAV communication technologies, this paper reveals that the existing standards of communications cannot meet the increasing requirements of UAV data transmission. Subsequently, four representative CS techniques, including compressed sensing, one-bit compressed sampling, phase retrieval and matrix completion, are briefly reviewed. Then, the simulations of compressed sensing and matrix completion technologies with real-world data are carried out to demonstrate the effectiveness which reveals that compressed sensing and matrix completion methods are able to significantly reduce the transmission time of data backhaul without changing the bandwidth. Ultimately, this paper also describes four research and development directions in the field of UAV data compression and recovery.

引言
1 UAV通信瓶颈
2 压缩采样技术
3 实验验证
4 展望:机会与挑战

图1 现有无线通信技术<br/>Fig.1 Existing wireless communication techniques

图1 现有无线通信技术
Fig.1 Existing wireless communication techniques

表1 UAV通信技术核心参数<br/>Table 1 Core parameters of UAV communication technologies

表1 UAV通信技术核心参数
Table 1 Core parameters of UAV communication technologies

图2 四种压缩采样技术在UAV数据传输中的作用<br/>Fig.2 (Color online)The role of four compression sampling techniques in UAV data transmission

图2 四种压缩采样技术在UAV数据传输中的作用
Fig.2 (Color online)The role of four compression sampling techniques in UAV data transmission

图3 实验流程图<br/>Fig.3 (Color online)Flowchart of experiment

图3 实验流程图
Fig.3 (Color online)Flowchart of experiment

图4 基于深度神经网络的压缩感知框架<br/>Fig.4 (Color online)CS framework based

图4 基于深度神经网络的压缩感知框架
Fig.4 (Color online)CS framework based

图5 基于原始数据和压缩感知数据的SAR成像<br/>Fig.5 SAR imaging based on original and CS data

图5 基于原始数据和压缩感知数据的SAR成像
Fig.5 SAR imaging based on original and CS data

图6 基于原始数据和矩阵补全数据的SAR成像<br/>Fig.6 SAR imaging based on original and MC data

图6 基于原始数据和矩阵补全数据的SAR成像
Fig.6 SAR imaging based on original and MC data

[1] SAMAD A M, KAMARULZAMAN N, HAMDANI M A, et al. The potential of unmanned aerial vehicle(UAV)for civilian and mapping application[C]// The 3rd Inter-national Conference on System Engineering and Technology. Shah Alam, Malaysia: IEEE, 2013: 313-318.
[2] STÖCKER C, BENNETT R, NEX F, et al. Review of the current state of UAV regulations[J]. Remote sensing, 2017, 9(5): 459.
[3] 全球无人机网.外军无人机数据链发展及趋势[EB/OL].[2019- 07-22]. https://www.81uav.cn/uav-news/ 197001/01/938.html
[4] DECUIR J. Introducing bluetooth smart: part 1: a look at both classicand new technologies[J]. IEEE Consumer Electronics Magazine, 2014, 3(1): 12-18.
[5] DECUIR J. Introducing bluetooth smart: part II: applications andupdates[J]. IEEE Consumer Electronics Magazine, 2014, 3(2): 25-29.
[6] BROWN T X, ARGROW B, DIXON C, et al. Ad hoc UAV ground network(AUGNet)[C]// The 3rd “Unmanned Unlimited”Technical Conference. Chicago, USA: AIAA, 2004: 1-11.
[7] YANMAZ E, KUSCHNIG R, BETTSTETTER C. Achieving air-groundcommunications in 802.11 networks with three-dimensional aerialmobility[J]. Proceedings of IEEE Conference on Computer Communications(INFOCOM). Turin, Italy: IEEE,2013: 120-124.
[8] MIAO L, DJOUANI K, WYK B J, et al. Evaluation andenhancement of IEEE 802.11p standard: a survey[J]. Mobile Computing, 2012,1(1): 15-30.
[9] OMEROVIC S. WiMAX overview[D]. Ljubljana, Slovenia: University of Ljubljana, 2006.
[10] GRANBOHM H, WIKLUND H. GPRS: general packet radio service[J]. Ericsson Review, 1999, 76(2): 82-88.
[11] HUBER J F, WEILER D, BRAND H. UMTS, the mobile multimediavision for IMT 2000: a focus on standardization[J]. IEEE CommunicationsMagazine, 2000, 38(9): 129-136.
[12] MARTÍN-SACRISTÁN D, MONSERRAT J F, CABREJAS-PENUELAS J, et al. On the way towards fourth-generationmobile: 3GPP LTE and LTE-advanced[J]. EURASIP Journal on WirelessCommunications and Network, 2009(1):354089.
[13] SORNINN. LoRa WAN 1.1 specification[EB/OL].(2017-10-11).https://lora-alliance.org/sites/default/files/2018- 04/lorawantm_specification_-v1.1.pdf
[14] CANDES E J, WAKIN M B. A introduction to compressive sampling[J]. IEEE Signal Processing Magazine, 2008, 25(2): 21-30.
[15] DONOHO D L. Compressed sensing[J]. IEEE Transaction on Information Theory, 2006, 52(4): 1289-1036.
[16] MALLAT S G, ZHANG Zhifeng. Matching pursuits with time-frequencydictionaries[J]. IEEE Transactions on signal processing, 1993, 41(12): 3397-3415.
[17] BLUMENSATH T, DAVIES M E. Iterative hard thresholding for compressed sensing[J]. Applied and Computational Harmonic Analysis, 2009, 27(3): 265-274.
[18] NEEDELL D, TROPP J A. CoSaMP: iterative signal recovery from incomplete and inaccurate samples[J]. Applied and communicational harmonic analysis, 2010, 53(12): 93-100.
[19] XIAO Yunhai, YANG Junfeng, YUAN Xiaoming. Alternating algorithms for total variation image reconstruction from random projections[J]. Inverse Problems and Imaging, 2012, 6(3): 547-563.
[20] DONG Weisheng, SHI Guangming, LI Xin, et al. Image reconstruction with locally adaptive sparsity and nonlocal robust regularization[J]. Signal Processing: Image Communication, 2012, 27(10): 1109-1122.
[21] HEGDE C, INDYK P, SCHMIDT L. A fast approximation algorithm for tree-sparse recovery[C]// IEEE Inter-national Symposium on Information. Honolulu, USA: IEEE, 2014: 1842-1846.
[22] MOUSAVI A, PATEL A B,BARANIUK R G. A deep learning approach to structured signal recovery[C]// The 53rd Annual Allerton Conference on Communication, Control, and Computing(Allerton). Monticello, USA: IEEE, 2015: 1336-1343.
[23] KULKARNI K, LOHIT S, TURAGA P, et al. Reconnet: non-iterative reconstruction of images from compressively sensed random measurements[C]// IEEE Conference on Computer Vison and Pattern Recognition. Las Vegas, UAS: IEEE, 2016: 449- 458.
[24] ZHANG Jian,GHANEM B. ISTA-Net: iterative shrinkage-thresholding algorithm inspireddeep network for image compressive sensing[EB/OL].(2017- 06-24)[2018- 06-18]. https://arxiv.org/abs/1706.07929v1
[25] DONG Xiao, ZHANG Yunhua. A map approach for 1-bit compressive sensing in synthetic aperture radar imaging[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(6): 1237-1241.
[26] 周崇彬. 单比特合成孔径雷达稀疏成像技术的研究[D].合肥:中国科学技术大学, 2016.
[27] BOUFOUNOS P T, BARANIUK R G. 1-bit compressive sensing[C]// The 42nd Annual Conference on Infor-mation Sciences and Systems. Princeton, USA: IEEE, 2008: 16-21.
[28] KARAHANOGLU N B, ERDOGAN H. Compressed sensing signal recovery via forward-backward pursuit[J]. Digital Signal Processing, 2013, 23(5): 1539-1548.
[29] PLAN Y, VERSHYNIN R. One-bit compressed sensing by linear programming[J]. Communications on Pure & Applied Mathematics, 2011, 66(8): 1275-1297.
[30] KNUDSON K, SAAB R, WARD R. One-bit compressive sensing with norm estimation[J]. IEEE Transactions on Information Theory, 2006, 62(5): 2748-2758.
[31] MIAO J, ISHIKAWA T, JOHNSON B, et al. High resolution 3D X-ray diffraction microscopy[J]. Physical Reviewer Letter, 2002, 89(8): 088303.
[32] PATTON L K, RIGLING B D. Phase retrieval for radar waveform optimization[J]. IEEE Transactions on Aerospace and Electronic System, 2012, 48(4): 3287-3302.
[33] SHECGTMAN Y, ELDAR Y C, COHEN O, et al. Phase retrieval with application to optical imaging: a contemporary overview[J]. IEEE Signal Processing Magazine, 2015, 32(3): 87-109.
[34] QIU T, BABU P, PALOMAR D P T. PRIME: phase retrieval via Majorization-minimization[J]. IEEE Transactions Signal Processing, 2016, 64(19): 5174-5186.
[35] QIU T, BABU P, PALOMAR D P T. Under-sampled sparse phase retrieval via majorization-minimization[J]. IEEE Transactions Signal Processing, 2017, 65(22): 5957-5969.
[36] LI Xiaopeng, HUANG Lei, SO H C, et al. A survey on matrix completion: perspective of signal processing[EB/OL].(2019- 01-25)[2019- 05- 07]. https://arxiv.org/abs/1901.10885.
[37] CANDÈS E J, RECHT B. Exact matrix completion via convexoptimization[J]. Foundations of Computational mathematics, 2009, 9(6): 717.
[38] CANDÈS E J, PLAN Y. Matrix completion with noise[J]. Proceedings of the IEEE, 2010, 98(6):925-936.
[39] FAZEL M. Matrix rank minimization with applications[D]. Stanford: Stanford University, 2002.
[40] CAI Jianfeng, CANDÈS E J,SHEN Zuowei. A singular value thresholdingalgorithm for matrix completion[J]. SIAM Journal Optimization. 2010, 20(4): 1956-1982.
[41] MA Shiqian, GOLDFARB D, CHEN Lifeng. Fixed point and Bregman iterative methods for matrix rank minimization[J]. Mathematical Programming, 2011, 128(1/2): 321-353.
[42] KESHAVAN R H, MONTANARI A, OH S. Matrix completion from a few entries[J]. IEEE Transactions on Information Theory, 2010, 56(6): 2980-2998.
[43] JIANG Xue, ZHONG Zhimeng, LIU Xingzhao, et al. Robust matrix completion via alternating projection[J]. IEEE Signal Processing Letters, 2017, 24(5): 579-583.

备注

引言

1 UAV通信瓶颈

2 压缩采样技术

3 实验验证

4 展望:机会与挑战

期刊信息