A new sensor exploits the fact that optimized radio waves can penetrate nonmetallic materials. The Motion Detection Radar (mdr) adds a new dimension to surveillance. Mdr can be effective even through wooden doors and concrete walls.
A new sensor exploits the fact that optimized radio waves can penetrate nonmetallic materials. The Motion Detection Radar (mdr) adds a new dimension to surveillance. Mdr can be effective even through wooden doors and concrete walls.
A new sensor exploits the fact that optimized radio waves can penetrate nonmetallic materials. The Motion Detection Radar (mdr) adds a new dimension to surveillance. Mdr can be effective even through wooden doors and concrete walls.
dazzling array of high tech solutions for military prob- lems. While typically effective for long range mass destruction, these solutions generally are not useful in combatting civilian crime. Our goal is to convert high tech Doll) capa- bilities into cost effective tools that help law enforcement agencies. For example, a new sensor has been designed based on technology developed for missile warhead fusing. This small, light weight, low power radar exploits the fact that opti- mized radio waves can penetrate non- metallic materials. This new surveillance capability can help pro- vide information about what is in a wall, ceiling, floor, on the other side of a door or even a concrete wall. generally line-of-sight (LOS) devices. They must have a clear visual view of the surveillance area. This obviously places a major limitation on police field operations. It is often difficult to set up this type of surveillance without being detected. Whats more, visual surveillance, even with a TV, can be tedious and hstraling. A person watching a TV screen showing an area where nothing is going on soon loses interest. As a result, other distractions or drowsi- ness can cause important activities to be missed. A signal, such as a tone or flashing light, that operates only when there is activity would help. The Motion Detection Radar (MDR) adds a new dimension to sur- veillance. MDR can be effective even through wooden doors and concrete walls, It requires very little setup and P-; Remote Monitor 4 I I Body Dielectric Concrete Constant - 80 Wall Dielectric Constant - 8 -12 . The Human Body Contains -65% Water . Water Has A High Dielectric Constant - Walls Have A Low Dielectric Constant ~ Fig. 1 Wd / s are opaque, the body i s a good reflector. Law enforcement agency surveil- lance typically incl udea; television (TV) cameras, infra-red (IR) sensors and hidden microphones, often with remote transmitters. These sensors have greatly enhanced the surveillance capabilities of the law enforcement agencies. However, these sensors are does not have any extemd wires con- nected to it. Unlike TV cameras and IR sensors that must be concealed, the MDR can be placed in a safe loca- hidden in a container with non-metal- lic walls. The MDR provides an effective and timely alert even if the tion on the other side of a wall or person doing the surveillance is pre- occupied, distracted or inadvertently nods off due to fatigue. The basic premise is that radio waves will penetrate most non-metal- lic materials. However, a number of factors must be considered to properly exploit this phmorriena. The cornpo- sition and thickness of the material to be penetrated is a prime factor in the initial design. Likewise, the reflectivi- ty and uniqueness of the actual targets on the other side of the wall must also be considered. Radio waves both pen- etrate and reflect off of surfaces of a non-metallic material. The surface roughness, dielectric constant and angle of incidence all affect the pene- tration characteristics. In Fig. 1, the radar waves are pene- trating a dry concrete wall with a dielectric constant between 8 and 12. The primary target on the other side of the wall is a moving person. Because the human body has a high concentration of water, which has a dielectric constant of 80 (distilled), the expected reflectivity is quite rea- sonable compared to other objects in view. While other metallic objects have greater reflectance, the radars ability to scnse motion makes the moving target unique and relatively easy to detect in this highly cluttered environment. The effects of velocity change, ubsorption, refraction and reflection all must be considered as shown in Fig. 2. The radio waves velocity is slowed by the square root of the dielectric constant of the non-metallic materials (Fig. 2a). A material with a dielectric constant of 4 would decrease the velocity of the radio wave by a factor of 2. Thus. the dielectric constant and the thickness of the material determine how much the apparent range to the target is DECEMBER 97IJ ANUARY 98 0278-6648/97/$10.00 0 1997 IEEE 23 Transmission Speed Non-Metallic Material Slows Waves (4 Absorption The Wall Absorbs (c) Refraction Loss Lar e Particles Cause The Rajar Beam To Break Up Refraction Waves Bend Through Dielectrics (b) Reflection The Wall Reflects (4 Diffraction t Waves Diffract Around Edges Fig. 2 Considerations when RF i s penetrating non-metallic maferials. Fig. 3 Measured one way loss: 4- 140 GHz modified each time the signal goes through the material. For most thin materials, this delay is insignificant; but, for others, it can make a consider- able difference. The absorption of the radio energy, while passing through the material, is affected by the materials physical make up (Fig. 2c). If sufficiently large, con- ductive, dissipative particles (such as carbon) are in the material, then depending upon the signal frequency, ohmic attenuation may cause sufficient loss to make the system unusable. If the dielectric particle size of the structures material is large with respect to the sig- nals wavelength, there will be intemal reflections and refraction. These will distort the signal wave front and cause excessive attenuation (Fig. 2e). For example, concrete with large internal stone aggregate will have far greater loss at 10 GHz than the same thickness of concrete made with fine sand. The refraction of the wave passing through the material (Fig. 2b) is also a function of the dielectric constant and thickness. The effect of refraction is small for homogeneous materials with particle sizes much smaller than one quarter wavelength. For construction concrete, the effect can be significant at higher frequencies. The radar designer must consider the loss through the mate- rial (which should be small) compared to the normal two way radar range loss (one over range to thefourth power). Diffraction (Fig. 2f) is caused by radio waves striking an objects edge and producing a scattering of the radio waves. This effect is predominant when metal objects are inside the wall the radar beam is penetrating. The refection of the radio wave is also influenced by the angle of incidence between the wave and the wall, as well as the distance between the radar anten- na and the wall (Fig. 2d). When the antenna is pointed directly at the wall, the maximum direct return from the wall is received back at the radar. When the antenna beam is pointed at an angle to the wall, the reflection from the wall back to the radar rapidly decreases. Reflections off walls that go back to objects on the radars side of the wall can become quite strong. All large flat surfaces act as mirrors to the radar sig- nal. The consequences of this charac- teristic are difficult to predict. This can lead to significant angle errors in assessing the location of the moving person at certain frequencies with cer- tain types of walls. The effect is also modified by the texture of the walls surface: the smoother the wall, the greater the effective reflection and the smaller the penetration. A potential radar system limitation is the ratio of the power received off the close wall relative to the power received off a dis- tant moving target. The loss of radio frequency energy as the beam passes through materials varies greatly with different conditions. These conditions include: age, chemical as well as mechanical construction, and the amount of metallic contaminates. 24 IEEE POTENTIAL Figure 3 shows a comparison of the measured m e way losses versus fre- quency for various common wall and building materials. The concrete blocks were standard 6 inches thick by 6 inches high by 12 inches long blocks with two holes in the middle. This left about 1 and 1/2 inches of concrete on each side of the holes. A number of these blocks were stacked to prevent direct energy radiation around the wall. The same setup was used for the boards and other materials measured. There was no attempt to maintain uniform thickness for these tests; however, the distance between the antennas was constant. The concrete blocks presented the greatest loss of the materials tested. The initial tests concentrated on the effects at the higher frequencies. While there is a temptation to directly convert lhe one way loss to two way losses, this may lead to erroneous results. While the loss alone may be translatable, the angle of incidence and the distance from the tar- get to the wall can have a greater effect than just the one way loss factor. Com- mon window glass, for example, can cause large signal losses at certain offset angles. As stated previously, this is not necessarily predictable, but must be con- sidered in the system design and setup. To determine the possibilities for practical radar operation through con- crete, tests were conducted at the lower frequencies, Figure 4a shows the one way loss measurements through com- mon 8 inch, pre-stressed, reinforced concrete walls for both horizontal (H) and vertical (V) antenna orientations. The typical one way loss of only 2 to 4 dB at 900 MHz was quite acceptable for radar penetration. Figure 4b shows the one way H-H loss measurements through a three foot thick reinforced concrete wall in the fre- quency range of 500 MHz to 2500 MHz. At 900 MHz, the loss increased to 6 to 8 dB, which was still quite accept- able. A number of different measure- ments at different locations along the wall were taken, while keeping the transmit and receive antennas of the bi- static radar at a constant 12 inch dis- tance from the wall. The losses were quite consistent when the antennas were both placed horizontally. Figure 4c shows the variation in attenuation when the antennas were placed vertically. Some of the variation in the data was caused by the vertical steel reinforcing bars mounted every 12 inches in the concrete. The distance of the receive antenna from the wall also caused wide variations in the one way received energy levels. This also may have been caused by the reinforcing steel, or by the interference of reflec- tions between the wall and the antenna. ix The Motion Detection Radar, shown in Fig. 5 (pg. 26), is contained in a high impact carrying case. The antenna is a flat plate 13 inches x 13 inches (33 cm x 33 cm) located in the lid of the case. The antenna radiates a +/- 45 degree conical beam out of the case's right side. The transmitter and receiver mod- ules are mounted on the antenna. They also fit into the lid of the carrying case. The control unit is visible in the left side oi the case. Two high current recharge- able NI-CAD (nickel-cadmium) batter- ies are located under the foam partitions. The VHF radio transmitter is mount- ed in the left front of the case (not visi- ble). I t can be removed for remote operation or it can be replaced by the audio amplifier shown below the case. The two remote VHF radio receivers are used to receive the target detection tones that are generated by the radar. The receivers can pick up the tones up to a mile from the radar. The three bat- tery chargers are shown just below the receivers, A 50 foot extension cable allows the antenna, and transmitter and receiver units to be remote up to that distance from the carrying case. The highly sensitive continuous wave (CW) phase detection radar has been approved by the Federal Commu- nications Commission (FCC) for opera- tion in the 902 to 928 MHz frequency band. There are restrictions imposed by the FCC which limit the power radiated from the antenna and signal harmonic content of any commercially sold prod- uct using this frequency band. While this is not necessarily the best frequency for material penetration, it is a reason- able compromise. This i q hecauqe it can be sold and used commercially, and has proven to be effective. A block diagram of the system is shown in Fig. 6. There are two different W-kl & V-V Qne Way Loss Through 8" Concrete Wall (a) H4-I bne Way Loss Through 3' Concrete (b} I I I I V-v one W& toss TfIrough 3 Concrete (c) %Lo 700 so0 400 400 liao ,;Do 400 2100 A00 *&lo Freauencv MHz " =Receive Antenna from W Fig. 4 One way loss through 8 inch and 36 inch reinforced concrete DECEMBER '97IJ ANUARY '98 25 Fig. 5 The highly porfable MDR antenna types used with this system. One is a high gain directional anten- na with approximately 9 dB gain. It produces a cone shaped pattern that is +/- 45 degrees wide at the -3 dB point. This antenna radiates from the right edge of the thin, 13 inch (33 cm) square antenna. It has a front to back ratio of better than 10:1. The optional antenna is a very broad beam omni-directional antenna. The omni-antenna is round, 114inches (.63 cm>in diameter, 18 inches (45.75 cm) long and has a gain of one. The omni- antenna provides large volume cover- age in an enclosed region, such as a room. Although i t provides much shorter range coverage than the high gain antenna, it is aspect independent. It can be placed into a small hole or dropped into an isolated location to observe motion. A transmitter and receiver (TR) unit is mounted on each antenna. A small cable for power and signal is connected from the radar control unit to the TR unit. The length of this cable can be extended to 100 feet (30.5 meters) or more for remote antenna operation, or when it is used in throw phones. (These are ruggedized telephones that can be thrown through a window or door to encourage communications with an uncooperative occupant.) The signal processor restricts the sensitivity of the radar to motions between 0.2 feet per second to 5 feet per second. This covers the range of motion that could be expected from a human being under most conditions. There are two outputs from the radar. The first is a relay closure which acti- vates anytime there is motion detected above the adjustable threshold setting. The second is an audio tone which varies in pitch in proportion to the rate of motion being detected. The pilot tone is 60 Hz when there is no motion. The tone rises to approximately 300 Hz when motion is detected. The tone will fluctuate in pitch with the motion. This gives a relative indication of the distance to the radar and the persons speed. A hand-held portable, commercial band (VHF) radio is controlled by the MDR relay. When motion is detected, the radio transmitter is keyed. The MDR audio tone is applied to the mike jack of the radio. One or more similar portable radio receivers are used to pickup the transmitted signal at ranges of up to one mile from the MDR. The development of the NLDR system has been completely an in-house com- mercial activity. Other hardware using similar technology has been developed for several US government organiza- tions to meet their unique needs. A number of three dimensional imaging radars have been delivered to the US government. These can produce a three dimensional (3D) image of a person or object in the field of view with better than 2 inch three dimension- al image resolution. The very portable 3D system has been used in a number of field operations while operating with portable power. A modification of the 3D imaging radar is the two dimensional (2D) sys- tem. This system can provide range and angle to targets through concrete walls with better than 6 inch range resolution. The 2D system uses the same signal processors and display as the 3D system. It can be installed on the 3D system in a few minutes. Because of the frequencies and bandwidths required for these more sophisticated systems, it is not possible to obtain FCC approval for commercial applications under the present regula- tions However, the US government is m&ng them available to local enforce- ment agencies with prior approval. Applications for ground penetration are being tested and evaluated The key limitation has been the non-unique nature of buried materials and the fact they are not moving It is difficult to tell the difference between a mine and a dead tree limb, or pop-bottle, just below the surface Both can be detected by radar, but there are not enough unique features to clearly identify the objects. We have been researching techniques that will greatly enhance both the range and the angle accuracies. The goal is to obtain enough resolution to automati- cally extract salient unique features of objects below the ground. Frazier, L. M., Surveillance Through Walls and Other Opaque Materials, presented at the IEEE 1996 National Radar Conference, Ann Arbor Michi- gan, 13-16 May 1996. Mr Frazier has been with Hughes/ General Dynamics for 40 years His experience includes design and field testing of many different types of radar and signal processing systems. This includes both pulsed and continuous wave radar systems for mono-static and bi-static radar applications. Over the past six years, he has developed a field portable, high resolution three- dimensional imaging radar which pro- vides isometric images of RF reflective targets. He has developed the hand held, motion detection, concrete pene- trating radar that fits in a briefcase. He is presently working on new concepts for ground and wall penetration and mapping radars. Fig. 6 Motion detection radar block diagram 26 IEEE POTENTIALS