Research Article | Open Access
Imaging Method Based on Time Reversal Channel Compensation
The conventional time reversal imaging (TRI) method builds imaging function by using the maximal value of signal amplitude. In this circumstance, some remote targets are missed (near-far problem) or low resolution is obtained in lossy and/or dispersive media, and too many transceivers are employed to locate targets, which increases the complexity and cost of system. To solve these problems, a novel TRI algorithm is presented in this paper. In order to achieve a high resolution, the signal amplitude corresponding to focal time observed at target position is used to reconstruct the target image. For disposing near-far problem and suppressing spurious images, combining with cross-correlation property and amplitude compensation, channel compensation function (CCF) is introduced. Moreover, the complexity and cost of system are reduced by employing only five transceivers to detect four targets whose number is close to that of transceivers. For the sake of demonstrating the practicability of the proposed analytical framework, the numerical experiments are actualized in both nondispersive-lossless (NDL) media and dispersive-conductive (DPC) media. Results show that the performance of the proposed method is superior to that of conventional TRI algorithm even under few echo signals.
In recent years, time reversal imaging (TRI) has attracted more attention because of its potential for target detection , medical imaging [2, 3], nondestructive detection , and communications [5, 6]. The basic procedure of TRI contains two stages: the scatterers generate/reflect a signal and a set of receivers record the echo signals in TR forward propagation (TRFP) stage. In TR backpropagation (TRBP) stage, these recorded signals are time reversed (e.g., using analog-to-digital (A/D) converters, digital-to-analog (D/A) converters, and computational TRI ). After that, the TR signals are propagated from receivers to scatterer(s) and then from scatterer(s) back to transmitters. When spatial reciprocity principle holds, rebroadcast wavefields will focus at the location(s) of the original scatterer(s).
These studies assume propagation in lossless media. However, in many applications, such as optical transmission systems , ultrawideband communications , and imaging systems [10–12], losses and/or dispersion are existent in the intervening media. As a result, TR invariance is broken and the conventional TRI operation cannot be directly applied. Therefore, compensation techniques are proposed in these applications. In [10, 11], a frequency dispersion compensation method based on conventional TRI algorithm is proposed, which consider cumulative dispersion effects stronger at later times and time-window the received signals for convenient. Combining with time reversal multiple signal classification (TR-MUSIC) algorithm, this method has also been adopted in ultrasound areas . However, adopting time-windowing may introduce some contaminated information, such as interference from artifacts, and overlapping factor has to be considered, which would affect imaging results seriously. Besides, conventional TRI method generates imaging map by using the maximal value of signal amplitude. Consider the propagation loss of electromagnetic (EM) wave and targets near transceivers get more power than that far from transceivers, which leads to remote targets undistinguished, that is, near-far problem , while TR-MUSIC needs to measure echo signals for constructing the multistatic data matrix (MDM) whose size is , where is the total number of transceivers [3, 13, 14]. Thus, the use of conventional TRI and TR-MUSIC algorithms will increase the complexity and cost of system. Moreover, only low frequency (MHz) is considered in these researches.
For solving these problems discussed above, a novel TRI method is proposed in this paper. The signal amplitude corresponding to focal time () observed at target position is used to reconstruct the target image. To our knowledge, this proposed imaging function is novel comparing with existing researches. By using this novel imaging function, we can not only obtain high resolution, but also avoid considering overlapping factor. As to the near-far problem and interference from artifacts, channel compensation function (CCF) is introduced to suppress them. CCF contains amplitude compensation (solving near-far problem) and cross-correlation property (alleviating interference from artifacts). Furthermore, in order to reduce the complexity and cost of system, only five transceivers are utilized to locate four targets with high resolution. Generally, for locating targets accurately, the number of transceivers is much larger than that of targets [13, 15], and locating targets whose number is close to that of transceivers with high resolution appears in very few literatures. To prove the practicability of proposed method, two numerical experiments are executed in nondispersive-lossless (NDL) media and dispersive-conductive (DPC) media on the condition that the center frequency of incident pulse is 2.45 GHz (WLAN frequency band).
2. Description of the Proposed Method
An imaging system comprised of a linear array of transmitters and receivers is considered in this section and the th transmitter and th receiver are located at and , respectively. It is further assumed that each transmitter sequentially transmits probing pulse and each receiver collects the echo signals. The frequency-domain representation of incident signal at target (located at ) can be illustrated aswhere is the Fourier transform of and is the Green function which satisfies the reduced wave equation  (representing the “propagator” from location to location at frequency ),where is Laplace operator, is the wavenumber, and is the Dirac delta function.
Using to represent the propagation loss characteristic of EM wave in transmission space/medium, we can express the echo signal received by the th receiver aswhere is the object profile. To implement the numerical studies, suppose that the imaging region is discrete into grids and each grid has a reflection coefficient , . Then, we can get the as 
Since time reversing a signal in time domain is equivalent to taking complex conjugate in frequency domain, the TR version of (3) can be represented as
Considering all receivers resubmitting corresponding TR signals to an arbitrary point in TRBP stage, the signal received by the th transmitter can be derived as
According to spatial reciprocity principle, when , the signal received from target position is
It can be found that the TR signal at target position is a focused signal. However, this focusing phenomenon does not happen away from the target position. With enough transceivers, the signal amplitude at target position is much larger than that of off-target position, so that the target can be located accurately by using conventional TRI method. Whereas this is complex and uneconomic, thus the use of a small quantity of transceivers is an inevitable development trend. However, directly decreasing the amount of transceivers will result in obtaining fewer echo signals, by which the targets cannot be distinguished with conventional TRI algorithm. In order to solve this problem, a novel imaging map is built by using signal amplitude at . According to TR concept, the TR signal at target positions will focus synchronously, and, thus, is the same at all target positions, but the time corresponding to the maximal value of signal amplitude observed at off-target positions is different and not equal to . In other words, as to target position, the maximal value of signal amplitude appears at ; however, as to off-target position, the signal amplitude is smaller than the maximal value at . Therefore, compared with conventional TRI algorithm, this new imaging map (corresponding to (11)) can enhance target resolution even under few echo signal conditions.
Further, in order to solve the near-far problem and the interference from artifacts, a CCF () is introduced. Actually, the performance of the compensation depends on several factors, such as the amount of losses (e.g., the moisture level in the soil), the target locations (e.g., depth), and the frequency of probe impulse.
In lossy and/or dispersive media, the signal received by each receiver is only changed with a phase shift and an amplitude attenuation , which can be expressed aswhere is the propagation direction of EM wave, is the transmission distance of EM wave, is the attenuation factor, and is the phase factor. The phase aberration correction at target position is automatically compensated by the TR operation via phase conjugation. Taking into account the amplitude aberration, we can obtain the channel compensation function aswhere , , and are the inverse Fourier transform of , , and , respectively. represents the cross-correlation function between and . Comparing (6) with (7), it can be found that the received TR signal from target position is similar to , but this does not happen at off-target positions. Therefore, at target position is larger than that at off-target positions; namely, compared with off-target positions, the effect of CCF is more obvious at target position, which is able to be used to suppress ghost images effectively. By multiplying amplitude compensation factor , the CCF can solve near-far problem well.
The imaging formula can be formed as
Combining with iterative TRI method, is equal to the iteration number. If TR signals are only propagated from receivers to original observed area in TRBP stage at the last iteration,
3. Numerical Experiments and Discussions
In this section, NDL media and DPC media are studied in numerical experiments A and B, respectively. In both experiments, the size of observation area is 0.5 × 0.5 m2. The Gaussian modulated pulse of 0.5 ns duration centered at 2.45 GHz is used as imaging probe signal. Omnidirectional antennas operated at WLAN frequency band (2.4–2.485 GHz) were used as transceivers, and these numerical experiments were simulated by MATLAB software.
For proving this method can work under few echo signal conditions, only five transceivers are employed to detect four targets. The positions of targets are (0.1 m, 0.06 m), (0.2 m, 0.32 m), (0.36 m, 0.14 m), and (0.4 m, 0.4 m), respectively, and transceivers are situated at (0.05 m, 0 m), (0.15 m, 0 m), (0.25 m, 0 m), (0.35 m, 0 m), and (0.45 m, 0 m), respectively, as shown in Figure 1. In order to facilitate comparison, all echo signals will be processed by using the conventional TRI algorithm and the proposed algorithm.
Lorentz model for the complex permittivity function [18, 19] can be presented aswithwhere is the number of species of media, is the medium susceptibility, is the resonant frequency for the species, and is the corresponding damping factor and and are the static and infinite frequency permittivities, respectively.
3.1. Numerical Experiment A: In NDL Media
As to NDL media, and are equal to zero ().
From the imaging map obtained by conventional TRI method, as illustrated in Figure 2(a), it can be seen that targets cannot be distinguished at all in the condition of few echo signals, and near-far problem is serious. We can also find that, by using the proposed imaging map function, the resolution can be enhanced obviously, but near-far problem still exists as shown in Figure 2(b). Further, combining with CCF, targets can be distinguished and we can obtain cleaner, higher-resolution image results even under few echo signal conditions as portrayed in Figure 3.
Moreover, the received echo signal strength from each target from being large to small is , while the value of the CCF corresponding to each target is opposite (). It can be observed that, compared with and , is located closer to transceivers, which leads to a small compensation, and, compared with , the echo signal from is smaller; thus, is darker than others in the imaging result.
3.2. Numerical Experiment B: In DPC Media
Take the clay loam with moisture level of 5%, for example. Due to larger loss and more complex dispersion in DPC media, the near-far problem is more serious, imaging result obtained by using conventional TRI algorithm is worse than that in NDL media, as shown in Figure 4(a), three remote targets disappear, and one near target is not able to be identified.
Figure 4(b) gives the imaging result obtained by using the proposed imaging map function, although resolution is enhanced, targets cannot be distinguished clearly, and near-far problem still exists as discussed in numerical experiment A; moreover, artifacts appear near the transceivers.
Therefore, in order to solve problem above, CCF is introduced and Figure 5 illustrates the corresponding imaging result. It can be seen that four targets can be located accurately with high-resolution compared with Figure 4, and artifacts disappear. The similar results can also be obtained by taking the clay loam with moisture level of 2.5% and 10%, for example.
In this work, according to propagation loss and/or dispersion of EM wave in transmission space/medium, we introduced CCF to compensate channel attenuation and suppress spurious images; it consists of amplitude compensation and cross-correlation property between the conjugate of incident pulse and received TR signal. The use of amplitude compensation can solve near-far problem well, and spurious images can be suppressed effectively by utilizing cross-correlation function. In addition, the resolution can be enhanced by using new imaging map, which is constructed by the signal amplitude corresponding to focal time observed at target position. Two numerical experiments are carried out to prove the practicability of the proposed algorithm. It is worth mentioning that targets whose number is close to that of transceivers can be located accurately even under few echo signal conditions.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
This work was supported by the NSFC-Guangdong under Grant no. U1035002, the University-Industry Key Project of the Department of Education of Guangdong Province under Grant no. CGZHZD1102, the Special Foundation for IOT in China under Grant no. 351, and the Major Special Foundation of Guangzhou under Grant no. 2014Y2-00218.
- Y. Jin, J. M. F. Moura, Y. Jiang, D. D. Stancil, and A. G. Cepni, “Time reversal detection in clutter: additional experimental results,” IEEE Transactions on Aerospace and Electronic Systems, vol. 47, no. 1, pp. 140–154, 2011.
- Y. Chen, E. Gunawan, K. S. Low, S.-C. Wang, Y. Kim, and C. B. Soh, “Pulse design for time reversal method as applied to ultrawideband microwave breast cancer detection: a two-dimensional analysis,” IEEE Transactions on Antennas and Propagation, vol. 55, no. 1, pp. 194–204, 2007.
- M. D. Hossain, A. S. Mohan, and M. J. Abedin, “Beamspace time-reversal microwave imaging for breast cancer detection,” IEEE Antennas and Wireless Propagation Letters, vol. 12, pp. 241–244, 2013.
- C. Fan, M. Pan, F. Luo, and B. Drinkwater, “Multi-frequency time-reversal-based imaging for ultrasonic nondestructive evaluation using full matrix capture,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 61, no. 12, pp. 2067–2074, 2014.
- Z. Liu and T. C. Yang, “On overhead reduction in time-reversed OFDM underwater acoustic communications,” IEEE Journal of Oceanic Engineering, vol. 39, no. 4, pp. 788–800, 2014.
- M. Yoon and C. Lee, “A TR-MISI serial prefilter for robustness to ISI and noise in indoor wireless communication system,” IEEE Signal Processing Letters, vol. 21, no. 4, pp. 386–389, 2014.
- M. Scalerandi, M. Griffa, and P. A. Johnson, “Robustness of computational time reversal imaging in media with elastic constant uncertainties,” Journal of Applied Physics, vol. 106, no. 11, Article ID 114911, 2009.
- S. Kumar, “Compensation of third-order dispersion using time reversal in optical transmission systems,” Optics Letters, vol. 32, no. 4, pp. 346–348, 2007.
- A. Dezfooliyan and A. M. Weiner, “Phase compensation communication technique against time reversal for ultra-wideband channels,” IET Communications, vol. 7, no. 12, pp. 1287–1295, 2013.
- M. E. Yavuz and F. L. Teixeira, “Frequency dispersion compensation in time reversal techniques for UWB electromagnetic waves,” IEEE Geoscience and Remote Sensing Letters, vol. 2, no. 2, pp. 233–237, 2005.
- M. E. Yavuz and F. L. Teixeira, “Compensation of frequency dispersion in time reversal techniques for ultra-wideband electromagnetic waves,” in Proceedings of the IEEE Antennas and Propagation Society International Symposium, vol. 1B, pp. 714–717, IEEE, Washington, DC, USA, July 2005.
- Y. Labyed and L. Huang, “Ultrasound time-reversal MUSIC imaging with diffraction and attenuation compensation,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 59, no. 10, pp. 2186–2200, 2012.
- G. Shi and A. Nehorai, “A relationship between time-reversal imaging and maximum likelihood scattering estimation,” IEEE Transactions on Signal Processing, vol. 55, no. 9, pp. 4707–4711, 2007.
- X.-F. Liu, B.-Z. Wang, and J. L.-W. Li, “Transmitting-mode time reversal imaging using MUSIC algorithm for surveillance in wireless sensor network,” IEEE Transactions on Antennas and Propagation, vol. 60, no. 1, pp. 220–230, 2012.
- T.-H. Liao, P.-C. Hsieh, and F.-C. Chen, “Subwavelength target detection using ultrawideband time-reversal techniques with a multilayered dielectric slab,” IEEE Antennas and Wireless Propagation Letters, vol. 8, pp. 835–838, 2009.
- L. Borcea, G. Papanicolaou, and C. Tsogka, “Theory and applications of time reversal and interferometric imaging,” Inverse Problems, vol. 19, no. 6, pp. S139–S164, 2003.
- P. Zhang, X. J. Zhang, and G. Y. Fang, “Comparison of the imaging resolutions of time reversal and back-projection algorithms in EM inverse scattering,” IEEE Geoscience and Remote Sensing Letters, vol. 10, no. 2, pp. 357–361, 2013.
- M. E. Yavuz and F. L. Teixeira, “Full time-domain DORT for ultrawideband electromagnetic fields in dispersive, random inhomogeneous media,” IEEE Transactions on Antennas and Propagation, vol. 54, no. 8, pp. 2305–2315, 2006.
- F. L. Teixeira, W. C. Chew, M. Straka, M. L. Oristaglio, and T. Wang, “Finite-difference time-domain simulation of ground penetrating radar on dispersive, inhomogeneous, and conductive soils,” IEEE Transactions on Geoscience and Remote Sensing, vol. 36, no. 6, pp. 1928–1937, 1998.
Copyright © 2015 Bing Li and Bin-Jie Hu. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.