Radar signal processing algorithm and simulation of detection system

Abstract

With its unique array arrangement, the detection system radar has both space diversity gain and waveform diversity gain, and is currently recognized as a stealth target buster. The detection system radar is applied to a high-speed moving platform. Using distributed cooperative detection technology, non-coherent fusion detection based on signals can further improve the detection of stealthy targets. Aiming at the high-speed motion radar signal processing algorithm, this paper mainly studies the following three aspects: the first content is the analysis of the waveform characteristics: the basic principles and characteristics of the radar are explained; then the three orthogonal waveforms commonly used in the radar are introduced, including Stepwise frequency division chirp signal, quadrature phase coded signal and mixed-signal; the second content detects radar targets and analyzes the correlation between the scattering coefficients of different radar channels; for scenarios where the scattering coefficients between the channels are non-coherent Introduced two kinds of non-coherent fusion detectors based on generalized likelihood ratio algorithm: centralized detector and double threshold detector; the third content radar multi-target pairing is aimed at the problem of radar multi-target pairing with large inertial navigation error. A multi-target pairing algorithm that uses target delay information and combines the radar’s multi-channel information redundancy characteristics is presented. An expression for judging the correctness of target pairing is derived, and the target pairing steps are given. The relationship between the amount of algorithm operation and the number of radar stations and the number of targets is analyzed in conclusion.

Keywords

High-speed motion platform generalized likelihood ratio based orthogonal waveform non-coherent fusion detection

1 Introduction

Due to the urgent military needs, radar was widely used and developed rapidly during the Second World War. With the advancement of science and technology, radar is not only an indispensable part of modern warfare. As civilian radar, it also plays an active role in navigation, geodetic surveying, and life detection. It can be seen that social and military needs have accelerated the development and update of radar technology. However, the rapid development of electronic information technology and the emergence of various active and passive interferences have seriously threatened the living environment of radars. At the same time, the application of stealth technology has also brought severe challenges to radar detection systems. These are undoubtedly exposed. The shortcomings of traditional radar are weak in anti-jamming, anti-stealth, anti-radiation, and anti-reconnaissance.

With the development of aviation technology, the speed of aircraft in electronic warfare is getting faster and faster [1 –3]. The radar is combined with high-speed moving platforms. Each radar station adopts cooperative combat technology to achieve the sharing of target information and pass non-coherent fusion detection at the central station it can further improve the detection performance of stealth targets and overcome the limitation of single radar. This is also the development trend of the radar in future electronic warfare. However, the high speed of the aircraft makes it very maneuverable, which brings huge challenges to the search, detection, and tracking of the radar. At the same time, the “three-span” phenomenon is prone to occur, that is, the unit moves across the distance, the unit across the Doppler unit, and the unit across the beam. If the “three-span” is not compensated, the coherent accumulation gain cannot be obtained.

The detection system radar uses an inertial navigation system for target positioning, there will be a large positioning error, which will seriously affect the radar’s estimation accuracy of target parameters, and also poses a problem for signal level fusion detection in a multi-target environment [4 –6]. Besides, since the radars of the detection system work independently, there will be three major synchronization issues: time, frequency, and space. Asynchronous Spatio-temporal frequencies will affect the detection performance of the radar system. At present, research on distributed radars on low- and medium-speed platforms has achieved a lot of theoretical research results, but research on distributed radars on high-speed platforms is very limited [7]. Therefore, research on radar signal processing algorithms on high-speed moving platforms Tactical confrontation has important research significance.

This paper mainly studies the radar signal processing algorithm of the high-speed moving platform detection system, which is mainly divided into three parts: The first part mainly analyzes the waveform characteristics of several common orthogonal signals, including cross-fuzzy functions, auto-correlation, cross-correlation characteristics, and Low interception characteristics. The second part introduces two non-coherent fusion detectors of radar and analyzes the impact of space diversity gain on target detection performance. The third part studies the existence of multi-target joint detection of detection system radar with large inertial navigation error. Multi-objective matching problem, a new target matching algorithm is proposed, and expression for judging the correctness of target matching is derived [8 –10].

2 Radar signal

2.1 Main categories of orthogonal waveforms

(1) Step frequency division linear frequency modulation signal

The frequency classification signal can be understood as a type of signal whose frequency spectrum is staggered in the frequency domain by frequency modulation. A generalized expression of the frequency-classified chirp signal will be given below. $rect (x) = {\begin{matrix} 1, | x | ⩽ 1 \\ 0, \end{matrix}$ (1)

When m is arranged in order, when m = m is satisfied, the frequency classification signal at this time is called step frequency division chirp signal. $SM 1 (t) = u (t) 2 π (\int α + (M - 1) \int ν t)$ (2)

In theory, in order for the signal of equation (3) to satisfy the orthogonality of equation (2), $f Δ T P = N 0$ (3)

In the above formula, it is any positive integer. It can be seen that the larger the value, the smaller it can be. At this time, the total bandwidth can be made narrower, which is very beneficial to reducing the digital sampling frequency of the receiving system and reducing the signal position in the search state: $B \sum = B + (M - 1) f Δ$ (4)

(2) Quadrature phase coded signal

Orthogonal phase coded signals are an important form of radar signal processing algorithms and simulation studies of transmitted signals in detection systems [11 –13]. They are widely used due to their high bandwidth utilization [14], their ideal fuzzy functions with approximate pin shapes, and good compression performance. The orthogonality of the quadrature-phase coded signal is mainly reflected in the coding sequence, so the carrier frequency of each transmitted signal can be the same. For a phase-encoded pulse signal with a code length of K and a subcode width of Tz, it can be expressed as: $S m (t) = \frac{1}{\sqrt{x}} \sum_{k = 0}^{k - 1} e ν (t - kT) E$ (5)

Among them Sm (t) Is the transmitting signal of the transmitting station m, ψM, K Represents the k-th subcode phase of the transmission signal of the transmitting station m, K TZ It can be determined that the pulse width is equal to KTZ, V (t) It is a rectangular pulse signal, which is defined as follows: When the phases m and k take a finite number of discrete values, a commonly used orthogonal phase-encoded signal-orthogonal uniform discrete phase-encoded signal-is obtained. According to different phase value ranges, it can be divided into two-phase code, four-phase code, or multi-phase code quadrature signal.

2.2 Mixed-signal

(1) The fuzzy function of conventional phase-encoded signals is in the shape of a pin, but it is very sensitive to Doppler frequency shifts and has limited bandwidth. LFM signals can achieve a large time-bandwidth product and are easy to process, but they have distance-speed coupling characteristics. The mixed-signal of the two signals combined modulation can improve the characteristics of the phase-coded signal with poor Doppler tolerance and the speed-distance coupling of the LFM signal. It has the advantages of an ideal blur function and a large time-bandwidth product. In this section, the principles of the quadrature-phase coded-LFM mixed-signal and stepped frequency division LFM-phase coded mixed signals will be introduced. $\frac{\frac{1}{N} \sum_{i = 1}^{N} {(x_{i} - r)}^{3}}{{(\frac{1}{N} \sum_{i = 1}^{N} {(x_{i} - r)}^{2})}^{\frac{3}{2}}}$ (6) $\Rightarrow 2 γ con = C_{x (2 mn)}^{- 1} (1 - P A FA)$ (7)

LFM signal and quadrature-phase coded pulse signals are convolved to obtain a new signal form-quadrature-phase coded-LFM mixed-signal [15]. A schematic diagram of the transmission waveform of the m-th transmitting station is given. The pulse signal is evenly divided into K sub-pulses. The time width of each sub-pulse is the symbol width, the phase is the phase of the coded signal, and the pulse is the LFM signal, that is, the linear frequency modulation in the code and the phase code between codes. $α = μ τ (m - n) f Δ fd$ (8)

(2) Stepped frequency division LFM-phase coded mixed-signal

An inner product of the LFM signal and the phase-encoded signal can be used to obtain another mixed signal-a stepped-frequency division LFM-phase-encoded mixed-signal [16 –18]. The transmission waveform of the m-th transmitting station is shown. It can be seen from the figure that each pulse is evenly divided into multiple sub-pulses, the number of which is K. The phase of each sub-pulse is the phase of the encoded signal, where m, k is the phase of the k-th sub-code of the signal transmitted by the transmitting station m. Different from the quadrature-phase code-LFM mixed-signal introduced in the previous section, the entire pulse width contains only one LFM signal, that is, the step-frequency division LFM signal between pulses and the phase-encoded signal within the pulse. $\sum i \sum jp (i, j, d, θ) / (1 + | i - j |)$ (9)

(3) Simulation model of radar signal processing system

Military requirements have more stringent requirements for radar performance, radar systems are relatively complex, and cycles are getting shorter and shorter. However, in practice, due to the influence of various significant factors, computer simulation technology must be used to assist the research and development of the radar system. During the analysis of the radar system and the construction of the simulation model, the simulation of the signal processing system is more important and must be processed by the computer. After the radar signal is constructed, the effective information must be extracted from the mixed-signal through specific algorithms and methods [19 –21]. Under normal conditions, the signal is mainly used to the output I / Q signals using quadrature mixing. The related analog information is converted by an A / D converter and converted into a digital signal that can be operated. It is converted into a digital signal that can be calculated, and then the FFT Fourier transform is used to transform the character information into Chirp-Z transform to transform it into the frequency domain form. Finally, the ideal result is obtained by the analog operation.

2.3 Waveform characteristics analysis

In order to obtain the waveform diversity gain of the distributed MIMO radar, the transmission waveform needs to maintain good orthogonality. Because the cross-correlation of waveforms will affect waveform diversity, the analysis of three types of orthogonal signals with the help of cross-fuzzy functions will help distributed MIMO radars better achieve waveform diversity. First, the mutual ambiguity function of distributed MIMO radar is given [22 –24].

(1) Step frequency division $R_{A} = q_{GE} l / 2 - (M_{E}^{r} + M_{E}^{l} + M_{GE}^{r} - M_{GE}^{l}) / l$ (10) $T = \sum_{i, j} \frac{(i - μ_{i}) (j - μ_{j}) p (i, j)}{σ_{i} σ_{j}}$ (11) $F = \sum i \sum jp (i, j, d, θ)^{2}$ (12)

It can be seen from the above formula that the value of the self-fuzzy function of the frequency-division chirp signal is consistent.

Traditional radars have severe performance loss for low RCS fluctuation target detection. Using radar signal processing algorithms and simulation research techniques of the detection system, non-coherent fusion detection can reduce the target RCS requirements at the same distance and the same detection performance [25 –27]. This chapter first analyzes the correlation between the scattering coefficients of echo signals from different channels. For non-coherent scattering coefficient scenarios, two non-coherent fusion detectors are introduced: a centralized detector and a dual-threshold detector. The signal model and the corresponding detector structure are given. The detection probability and false alarm probability expressions of the centralized detector, double threshold detector are derived. The influence of spatial diversity gain on target detection performance is analyzed. Finally, the detection performance of two non-coherent fusion detectors is simulated and compared.

(2) Waveform characteristics analysis of cloth MIMO radar

Due to the large spacing of distributed radar antennas, the observation angles of the same target by different antennas are quite different. At this time, the echo signals of the transmitting and receiving channels can be considered independent and have spatial diversity gain. The MIMO radar uses waveform diversity technology, and matched signals can be used to separate the echo signals of different channels during the reception, which increases the degree of freedom of the system and has a waveform diversity gain. The distributed MIMO radar combines the advantages of both so that it has both spatial diversity gain and waveform diversity gain. In general, the degree of influence of waveform diversity technology on the detection performance of a system is mainly related to the orthogonality of the transmitted waveforms, and different waveforms will have different effects on the combat capability of the radar system. In addition, in electronic warfare, the waveforms emitted by radars need to have low interception characteristics. Therefore, the analysis of radar waveform characteristics is part of the basic work of radar system design.

(3) Target echo simulation

In the parametric simulation, the engineering simplification of echo simulation is not much, and some data still needs to be generated according to various models. The specific process is similar to the simulation of signal flow. But it is not waveform data that is generated, but parameters. Finally, the echo waveform is simulated based on these data. In the parameterization process, the parameters and format of the target can be set according to the requirements of the system, such as distance, azimuth, elevation, and amplitude. Clutter can be used as a target, and the format of the parameters is the same as the target. For deceptive interference, the format is the same, but the angle information should be omitted (depending on the position of the incoming beam), and the final azimuth and pitch of the fake target are the same as the main lobe of the antenna. For the interference signal, there may be many parameters [28], such as the distance of the jammer, the attitude of the jammer, the signal polarization method, etc. These must also be set according to the system requirements. Deception jamming generally delays the pulses received by the reconnaissance plane, and may also do some processing and add other deceptive methods. Then the delay is a change in distance. According to the delay, the repetition period can be modulo to convert the position of the false target in the repetition period, that is, the distance.

3 Radar signal experiment

3.1 Experimental objectives

For the high-speed motion detection system radar, the position information of each platform provided by the inertial navigation system will have large errors, which poses a problem for multi-target spatial pairing. Therefore, in this chapter, aiming at the problem of radar target pairing in the detection system with large inertial navigation error, this paper proposes a method using target delay information and multi-channel information redundancy to complete target pairing. Each antenna of this system can send and receive, and transmit signals orthogonal to each other. This chapter first introduces the problem of multi-channel and multi-objective pairing, and then introduces the principle of the algorithm in detail, derives expressions to determine the correctness of the target pairing, gives the corresponding target pairing steps, and simplifies the complexity of the algorithm Finally, the validity of the method was verified by Monte-Carlo simulation.

3.2 Experimental assumptions

For the sake of analysis, it is assumed that the distributed MIMO radar system is composed of 3 independent phased array radars. There are three targets in space. Each radar and its transmitted signal are matched and filtered. One path can be separated. The delay of 3 targets, a total of 3 such paths, respectively (τ¹, τ², τ³), (τ¹, τ², τ³), (τ¹, τ², τ³). The superscript indicates the target number, and the subscript indicates the radar number [29 –31].

3.3 Experimental steps

The above analysis is carried out under the condition that the number of radars is 3 and the number of targets is 3. In order to obtain a more general target matching step, it is assumed that the number of radars is M(M≥2), The number of targets is Q (Q≥2), M>Q, When the number of radars is greater than the number of targets; No matter how the path is selected, there will be a case where two paths have selected the same target, that is, there is no case where the target pairing fails 2.

Step 1. Set the threshold according to the formula ξ, Initial number of successful pairings t = 0, Choose one target for all paths, which are q₁, q₂, q_M, The subscript represents the own path number;

Step 2. Calculate that all distances corresponding to the selected target are similar to dmn, which is equal to dmn = dnm

Step 3. Based on the target delay measured on his path, similar to equation (7), calculate whether all the following inequalities hold, if they exist (i = 1, 2, Q). If the following formula is established, then the selected target is successfully paired, and then go to step 4, otherwise go to step 5.

Step 4. Delete all successfully paired targets on all paths. t = t + 1, If t = Q, End algorithms; otherwise, select a target from the remaining targets on path 1 and return to step 2. From this, you can determine the path that failed to match. Select a path that fails to match and select the target again and return to step 2. MQ, the number of radars is less than or equal to the number of targets.

Step 5. At this time, the target pairing fails to meet the situation 1, that is, there are both successful and failed paths.

At this time, the target pairing steps are:

Steps 1 to 4 are the same as the pairing steps in MQ;

Step 5. Discuss by case. If all the paths, that is, m, n satisfy the formula, then it meets the failure condition 2 (all the paths choose different targets), and choose one of the paths to reselect the target. 8) If it is true, then the failure condition 1 is met, the paired correct path is excluded, and a target path that fails to match is selected again, and the target is selected, and the process returns to step 2.

4 Radar signal simulation analysis

4.1 Simulation

As shown in Table 1, the correctness of the derived expressions to determine the correctness of the target is verified, and ranging errors and inertial navigation errors are introduced. Assume that the detection system radar system consists of 3 radars, and the number of targets is 3, that is, M3, Q3. The absolute value of the radar position error is evenly distributed within a range of 2 km. The actual coordinates and inertial guidance of each radar and target in space are given in Table 1.

Table 1
Radar and target coordinates

Coordinate Reality Internal conductor

Radar 1 (0,0,20) (0.143,1.845,18.817)

Radar 2 (50,0,30) (19.578,1.593,31.847)

Radar 3 (0,50,20) (1.962,11.100,31.269)

Target 1 (0,120,0)

Target 2 (2,125,0)

Target 3 (2,822,0)

Coordinate	Reality	Internal conductor
Radar 1	(0,0,20)	(0.143,1.845,18.817)
Radar 2	(50,0,30)	(19.578,1.593,31.847)
Radar 3	(0,50,20)	(1.962,11.100,31.269)
Target 1	(0,120,0)
Target 2	(2,125,0)
Target 3	(2,822,0)

The ranging error is related to the radar range resolution. It is assumed that the radar signal bandwidth is 30 MHz and the resolution is 5 m. It is also assumed that the slope distance error is evenly distributed in the range of 5 m. After introducing the ranging error, the radar system’s own path and target delay information on other paths separated by matching filtering are shown in Tables 1 and 2.

Table 2

Each target delay on its own path

Target delay(us)	Way 4	Way 5	Way 6
Target 1	779.80	645.42	650.86
Target 2	715.52	645.58	649.94
Target 3	730.72	658.67	665.14

First set the threshold. Calculate the small target distance difference according to the minimum target delay difference of 1 min, and the corresponding relationship is 1 min / 2.

As shown in Table 2, the minimum target delay difference on the own path is the difference between the two values in the circle, and thus 15 m can be obtained. Because of 0 min, the threshold can be set to 12 m here.

First verify the correctness of expression (6) when the target pairing is successful, where the data shown in Table 3 are the calculated values of formula (8). In the simulation, all targets have chosen target 1. At this time, equation (8) should be established. Observing Table 4, it can be found that all the values in the circle are less than the threshold of 12 m. It can be understood that paths 1, 2, 1, 3, and 2, 3 select the same target, that is, all paths select the same target, and the targets are successfully paired Corresponds to equation (8). Note that you only know that all paths choose the same target, and you do not know that you chose to target 1, but this does not affect target matching.

Table 3

Each target delay on its own path

Target delay(us)	Way 1	Way 2	Way 3
Target 1	751.0	421.9	579.8
Target 2	711.1	719.8	580.0
Target 3	724.3	737.1	693.1

Table 4

Successful pairing

Pair(km)	D₁₂(way 4)	D₁₃(way 5)	D₂₃(way 6)
Target 1	0.0037	0.0074	0.0072
Target 2	0.2853	0.0562	0.2846
Target 3	4.2420	3.9450	4.2671

Next, verify the correctness of the expressions (5)∼(7) and (8) in the two cases where the target pairing fails. First, verify that the pairing fails 1, that is, only two paths choose the same destination. The data shown in Table 4 are the calculated values of equations (3) to (5). In the simulation, path 1 and 2 select targets 1, and path 3 select target 2. At this time, (3)∼(5) should be satisfied. As shown in Table 5 and Fig. 1, it can be found that only the value in the circle is less than the threshold of 12 m, which means that only paths 1, 2 select the same target, and path 3 selects different targets, corresponding to equations (6)∼(7) correctness. Similarly, I don’t know that paths 1, 2 choose to target 1.

Table 5

Failure to match 1

Pair(km)	d₁₂(way 4)	d₁₃(way 5)	d₂₃(way 6)
Target 1	0	0.0186	0.0156
Target 2	0.272	0.0154	0.3046
Target 3	4.5670	3.9561	4.5752

Fig. 1

Failure to match 1.

Here, verification fails 2, that is, all the paths have chosen different targets. In the simulation, paths 1, 2, and 3 select targets 1, 2, and 3, respectively. The data in Table 6 are calculated values of Equation (7). Looking at Table 6, it can be found that all the data in the table are greater than the threshold of 12 m, which means that all paths choose different targets, and the corresponding formula (7) verifies the correctness of the expression. Note that only three paths are selected to different targets, and the specific selection is not known. It can be seen that no matter whether the pairing succeeds or fails, there will be corresponding expressions.

Table 6

Failure to match 2

Pair(km)	D₁₂(way 4)	D₁₃(way 5)	D₂₃(way 6)
Target 1	0.3079	1.9883	1.6745
Target 2	0.0257	1.9863	1.9680
Target 3	4.5546	1.9869	2.5841

4.2 Simulation simulation theory 2

(1) Monte Carlo simulation experiments verify the performance of the proposed algorithm, that is, the variation curve of the pairing accuracy with the inertial navigation error and the ranging error. The number of radars, targets, radar positions, target positions, and error distribution is the same as those in simulation 1. The number of Monte-Carlo simulation experiments is 1000. As shown in Fig. 2, the curve of the pairing rate and the slope distance error of a single target pairing with three different initial target selections is given. The three initial situations are:

Fig. 2

Map of the variation of pairing rate with the deviation error.

Case 1: The same target is selected on the path, and the initial pairing is successful;

Case 2: Only two of the paths choose the same target, and the initialization pairing fails-pairing failure 1;

Case 3: Different paths have selected different targets and initial pairing failed-pairing failed2.

(2) As shown in Fig. 2, in the three cases, the proposed method can complete multi-target pairing with a high matching rate. Under the condition of 5 m ranging error, the pairing rate in case 1 can reach 97.6%, while in case 3 (worst case), the pairing rate can reach 95.6%, although it has been reduced. Secondly, as the ranging error increases, the pairing rate will decrease. This is because when the slope distance error increases, the calculated value of equation (12) will become larger. With a fixed threshold, the result of Equation (8) is more likely to exceed the threshold when the pairing is successful. The situation is not consistent, and the matching rate is reduced.

Figure 3 uses Monte-Carlo simulation experiments to analyze the relationship between the pairing rate and inertial navigation error. Consider the case where the target is initially selected for case 3, and the fixed ranging error is 5 m (also subject to a uniform distribution in the range of 5 m). As shown in Fig. 3, as the inertial navigation error increases, the pairing rate is basically maintained at 95%. It can be seen that the proposed method is mainly affected by ranging errors, but is not sensitive to inertial navigation errors. This method greatly reduces the effect of large inertial navigation errors on the multi-target pairing process.

Fig. 3

Theoretical detection curve-non-undulating target model.

As shown in Fig. 3, in the three cases, the proposed method can complete a multi-target pairing with a high matching rate. Under the condition of 5 m ranging error, the matching rate of Case 1 can reach 97.6%, while the matching rate of Case 3 (worst case) is reduced, it can still reach 95.6%. Secondly, as the ranging error increases, the pairing rate will decrease. This is because when the slope distance error increases, the calculated value of the Equation (4–17) will become larger. With a fixed threshold, the result of Equation (4 –7) will more easily exceed the threshold when the pairing is successful, resulting in the expression result. The matching with the real target is not consistent, and the matching rate is reduced.

As shown in Fig. 4, the relationship between the pairing rate and the inertial navigation error is analyzed through Monte-Carlo simulation experiments. Consider the case where the target is initially selected for case 3, and the fixed ranging error is 5 m (also subject to a uniform distribution in the range of 5 m). It can be seen from Figs. 4–7 that as the inertial navigation error increases, the pairing rate is basically maintained at 95%. It can be seen that the proposed method is mainly affected by ranging errors, but is not sensitive to inertial navigation errors. This method greatly reduces the effect of large inertial navigation errors on the multi-target pairing process.

Fig. 4

Comparison of actual and theoretical detection curves.

5 Conclusions

In recent years, multiple-input multiple-output technology in communication systems has been widely used in the field of radar. According to the configuration of the antenna, it can be roughly divided into two categories: centralized MIMO radar and detection system radar. Because the radar of the detection system uses a large space configuration, the transmission waveforms are orthogonal to each other, so that it has both space diversity gain and waveform diversity gain. Combining detection system radar with high-speed motion platforms and adopting coordinated combat technology can greatly improve the combat capability of the system, which is the development direction of future modern wars.

Based on the basic principles of distributed MIMO radar, this article analyzes several common distributed MIMO radar quadrature waveforms, including stepped frequency division signal, code division signal, and mixed-signal, where the mixed-signal includes quadrature-phase coded-LFM mixed-signal and stepped frequency division LFM-phase coded mixed signal. Then the cross-fuzzy function is used to analyze the waveform characteristics of these three types of signals. When the bandwidth is limited, the mixed-signal can provide more orthogonal channels. Finally, the low interception characteristics of these three types of signals are discussed based on the low interception factor, and the conclusion that the anti-interception performance of the mixed signal is the best is drawn.

Due to the severe performance loss of traditional radars for low RCS target detection, the use of non-coherent fusion detection through distributed radar technology can reduce the target RCS requirements at the same distance and the same detection performance. This paper analyzes the correlation between the scattering coefficients of the echo signals of different channels and points out that the correlation is related to the radar station configuration, target position, target size, and carrier frequency. For non-coherent scattering coefficients, two non-coherent fusion detectors are introduced: a centralized detector and a dual-threshold detector. The signal model and detector structure are given. The detection probabilities and false alarm probability expressions of the centralized detector and the double threshold detector are derived, and simulations verify their correctness. The detection performance of the two detectors is compared. Under the condition of data rate constraints, the dual-threshold detector will lose a smaller signal-to-noise ratio than the centralized detector but still maintain good detection performance. In addition, the effect of spatial diversity gain on target detection performance is analyzed. Simulation results show that compared to single-channel detection, multi-channel non-coherent fusion detection can bring more improvement in signal-to-noise ratio, and fluctuates at the target RCS. Under the circumstances, the detection performance improvement effect obtained by the spatial diversity gain is more obvious.

Due to the large inertial navigation error of the radar of the high-speed moving platform detection system, the radar positioning error caused by the inertial navigation error will bring great difficulties to multi-target pairing. Aiming at this problem, this paper proposes a multi-target pairing algorithm suitable for large inertial navigation error scenarios. This algorithm uses the characteristics of target delay information and multi-channel information redundancy to complete target pairing. First, the radar multi-target pairing problem of the detection system is introduced, and the central idea and algorithm principle of the algorithm is described in combination with spatial geometry. The expressions for judging the correctness of target pairings are derived, and the corresponding expression results in different pairing situations are analyzed. Next, the steps of target pairing are given, and the algorithm complexity is analyzed. It is concluded that the calculation amount is only related to the number of radar stations and targets in the radar system of the detection system, and the exponential relationship between the number of radars and the calculation amount. Finally, Monte-Carlo simulation was used to verify the correctness of the derived expressions. The performance of the algorithm was analyzed. It was pointed out that the algorithm had nothing to do with radar positioning errors, but it failed when the detection targets were located in close range units.

Footnotes

Acknowledgments

This Work Was Supported by the National Natural Science Foundation of China (61773016, 61473222, 61873201).

References

Boccardi

, Heath

R.W.

and Lozano

, Five disruptive technology directions for 5G, Communications Magazine IEEE 52(2) (2015), 74–80.

Moo

P.W.

, Multiple-input multiple-output radar search strategies for high-velocity targets, Iet Radar Sonar & Navigation 5(3) (2017), 256–265.

Hassanien

, Vorobyov

S.A.

and Gershman

A.B.

, Moving target parameters estimation in Noncoherent MIMO Radar Systems, IEEE Transactions on Signal Processing 60(5) (2015), 2354–2361.

Shen

C.A.

, Yu

C.P.

and Huang

C.H.

, Algorithm, and Architecture of Configurable Joint Detection and Decoding for MIMO Wireless Communications with Convolutional Codes, IEEE Transactions on Very Large Scale Integration Systems 24(2) (2016), 587–599.

Fishler

, Haimovich

and Blum

, MIMO radar: an idea whose time has come, Radar Conference, Proceedings of the IEEE. IEEE Xplore 2015 (2015), 71–78.

Ying

, He

, Liu

, et al., Binary orthogonal code design for MIMO radar systems, International Symposium on Intelligent Signal Processing and Communication Systems. IEEE 2015, 1–4.

Han

, Li

, Su

, Guo

and Zhao

, Secrecy capacity optimization via cooperative relaying and jamming for WANETs, IEEE Transactions on Parallel and Distributed Systems 26(4) (2014), 1117–1128.

Masmoudi

and Le-Ngoc

, Subspace-based self-interference cancellation for full-duplex MIMO transceivers, Eurasip Journal on Wireless Communications & Networking 2017(1) (2017), 55.

Pelich

, Longépé

and Mercier

, AIS-Based Evaluation of Target Detectors and SAR Sensors Characteristics for Maritime Surveillance, IEEE Journal of Selected Topics in Applied Earth Observations & Remote Sensing 8(8) (2017), 3892–3901.

10.

Liang

, Cui

and Yu

, Analysis of Direction Finding Sensitivity of MIMO Radar with System Errors, Wireless Personal Communications 87(1) (2016), 1–21.

11.

Gogineni

and Nehorai

, Game-theoretic design for polarimetric MIMO radar target detection, Signal Processing 92(5) (2015), 1281–1289.

12.

Sheikhi

and Zamani

, Temporal coherent adaptive target detection for multi-input multi-output radars in clutter, Iet Radar Sonar & Navigation 2(2) (2018), 86–96.

13.

Fishler

, Haimovich

and Blum

R.S.

, Spatial diversity in radars-models and detection performance, IEEE Transactions on Signal Processing 54(3) (2016), 823–838.

14.

Tsai

S.B.

, Chien

M.F.

, Xue

, Li

, et al., Using the Fuzzy DEMATEL to determine Environmental Performance: A Case of Printed Circuit Board Industry in Taiwan, Plos One 10(6) (2015), e0129153.

15.

Saucedo-Jimenez

, Ortiz-López

, Garibay-Febles

and Palacios-González

, Sem and Hrtem Analysis of Carbon Nanostructures Synthesized with a New Pulsed Electric Arc Discharge Technique, Acta Microscopica 22(3) (2013), 289–299.

16.

Haimovich

A.M.

, Blum

R.S.

and Cimini

L.J.

, MIMO radar with widely separated antennas, IEEE Signal Processing Magazine 25(1) (2017), 116–129.

17.

Gao

, Li

and Himed

, Knowledge-Aided Range-Spread Target Detection for Distributed MIMO Radar in Non-Homogeneous Environments, IEEE Transactions on Signal Processing PP(99) (2016), 1-1.

18.

Godrich

, Haimovich

A.M.

and Blum

R.S.

, Target localisation techniques and tools for multiple-input multiple-output radar, Iet Radar Sonar Navigation 3(4) (2015), 314–327.

19.

Liang

, Leung

C.S.

and So

H.C.

, Lagrange programming neural network approach for target localization in distributed MIMO radar, IEEE Transactions on Signal Processing 64(6) (2016), 1574–1585.

20.

Cheng

, Liao

and He

, Spectrally Compatible Waveform Design for MIMO Radar in the Presence of Multiple Targets, IEEE Transactions on Signal Processing 66(13) (2018), 1–1.

21.

and Stoica

, MIMO radar with colocated antennas, IEEE Signal Processing Magazine 24(5) (2017), 106–114.

22.

Lehmann

N.H.

, Fishler

and Haimovich

A.M.

, Evaluation of transmits diversity in MIMO-radar direction finding, IEEE Transactions on Signal Processing 55(5) (2017), 2215–2225.

23.

Liang

, Cui

and Yu

, Analysis on Direction Finding Sensitivity of MIMO Radar with System Errors, Wireless Personal Communications 87(1) (2016), 1–21.

24.

Guruswamy

and Blum

, Ambiguity Optimization for Frequency-Hopping Waveforms in MIMO Radars with Arbitrary Antenna Separations, IEEE Signal Processing Letters 23(9) (2016), 1231–1235.

25.

Deligiannis

, Panoui

and Lambotharan

, Game-Theoretic Power Allocation and the Nash Equilibrium Analysis for a Multistatic MIMO Radar Network, IEEE Transactions on Signal Processing PP(99) (2017), 1-1.

26.

Yan

, Liu

and Pu

, Joint Beam Selection and Power Allocation for Multiple Target Tracking in Netted Colocated MIMO Radar System, IEEE Transactions on Signal Processing PP(99) (2016), 1-1

27.

Cheng

, He

and Liao

, MIMO radar waveform design with PAPR and similarity constraints, IEEE Transactions on Signal Processing PP(99) (2018), 1-1.

28.

, Zhang

, Liu

, Sun

, Cioffi

J.M.

and Yang

, Overhearing Protocol Design Exploiting Intercell Interference in Cooperative Green Networks, IEEE Transactions on Vehicular Technology 65(1) (2016), 441–446.

29.

, Liao

and Huang

, Robust adaptive beamforming for fast-moving target detection with FDA-STAP radar, IEEE Transactions on Signal Processing 65(4) (2016), 973–984.

30.

Quan

, Li

and Hu

, FM Sequence Optimization of Chaotic-Based Random Stepped Frequency Signal in Through-the-Wall Radar, Iet Signal Processing 11(7) (2017), 830–837.

31.

Kelly

E.J.

, An adaptive detection algorithm, IEEE Transactions on Aerospace & Electronic Systems AES-22(2) (2017), 115–127.

Radar signal processing algorithm and simulation of detection system

Abstract

Keywords

1 Introduction

2 Radar signal

2.1 Main categories of orthogonal waveforms

3.1 Experimental objectives

3.2 Experimental assumptions

3.3 Experimental steps

4 Radar signal simulation analysis

4.1 Simulation

Table 1 Radar and target coordinates Coordinate Reality Internal conductor Radar 1 (0,0,20) (0.143,1.845,18.817) Radar 2 (50,0,30) (19.578,1.593,31.847) Radar 3 (0,50,20) (1.962,11.100,31.269) Target 1 (0,120,0) Target 2 (2,125,0) Target 3 (2,822,0)

Footnotes

Acknowledgments

References

Table 1
Radar and target coordinates

Coordinate Reality Internal conductor

Radar 1 (0,0,20) (0.143,1.845,18.817)

Radar 2 (50,0,30) (19.578,1.593,31.847)

Radar 3 (0,50,20) (1.962,11.100,31.269)

Target 1 (0,120,0)

Target 2 (2,125,0)

Target 3 (2,822,0)