VISVESVARAYA TECHNOLOGICAL UNIVERSITY Jnana Sangama, Belgaum-590014, Karnataka.

EVALUATION OF POLYPHASE FFT ARCHITECTURE FOR PULSE DETECTION AND MEASUREMENTcarried out atDefence Avionics Research Establishment,DRDO

SUBMITTED IN PARTIAL FULFILLMENT FOR THE AWARD OF OfMASTER OF ENGINEERINGInDIGITAL ELECTRONICS AND COMMUNICATION ENGGFor the academic year 2013-2014 Submitted by JEEVITHA TUSN: 1DS12LEC06

Under the Guidance of

INTERNAL GUIDE NAME: EXTERNAL GUIDE NAME:Mrs. Kiran Gupta Mr.Hemanth Vasant Paranjape Professor, Scientist DDept of E & C DARE,DRDO

Department of Electronics & CommunicationDAYANANDA SAGAR COLLEGE OF ENGINEERINGShavige Malleshwara Hills, Kumaraswamy Layout, Bangalore 560 078

CERTIFICATE

This is to certify that Jeevitha T carried out the Project on Evaluation of polyphase FFT architecture for pulse detection and measurement under my guidance for the subject Project for the 3rd semester for the Master of Technology in Digital Electronics & Communication at DayanandaSagar College of Engineering.

Head of Department,Assistant Professor,Dr.Girish.V.AttimaradProf. Kiran Gupta

------------------------------------------------------------------

ACKNOWLEDGEMENT

The satisfaction and euphoria that accompany the successful completion of any task would be incomplete without the mention of the people who made it possible, whose constant guidance and encouragement crowned our effort with success.

I wish to place on record my grateful thanks to Dr.Girish.V.Attimarad,Head of the Department, Electronics and Communication Engineering, for providing encouragement and oppurtunity.

I extend my gratitude to Dr. K.L. Sudha for her kind co-ordination in the project phase.

I would like to thank my external project guide, Mr. Hemant Vasant Paranjape, Scientist D, Defence Avionics Research Establishment,DRDO for her valuable guidance and time to time evaluation of the Project. I would also like to thank Mr. Abhijit S Kulkarni, Scientist C, Defence Avionics Research Establishment,DRDO for his timely guidance and support.

I would like to thank my internal project guide, Prof.Kiran Gupta, Department of Electronics and Communication Engineering for her valuable guidance and time to time evaluation of the Project.

My thanks to the staff of the department of ECE, DSCE for the kind cooperation.

Submitted by:

JEEVITHA T1DS12LEC06

Evaluation of Polyphase Filter architecture for Pulse detection and measurement

Introduction:

Filtering is an important and necessary operation in any receiver system. A filter is a system that alters the spectral content of input signals in a certain way. Common objectives for filtering include improving signal quality, extracting signal information, and separating signal components. Filtering can be performed in the analog or digital domains. The major advantages of digital processing over analog processing are programmability, reproducibility, flexibility and stability. Since digital processing algorithms are implemented as computer programs or firmware, it is very easy to change any parameter (for example filter gain, filter pass band width etc.) compared to analog processing.

Abstract:

Polyphase filtering is a multirate signal processing operation that leads to an efficient filtering structure for hardware implementation. Polyphase filtering parallelizes the filtering operation through decimation of the filter coefficients, h(n). Polyphase filters can also be used to sub-band the frequency spectrum, thus producing a filter bank. FIR filters are commonly used in DSP implementations. FIR filters are linear phase filters, so phase distortion is avoided

Block Diagram:

Design Procedure:

The incoming N data samples are distributed in M branches and an M point FFT is performed. A given M point FFT divides the input frequency band into M fs/M filters. Number of points M is selected based on time resolution required. The decimation results in gaps in the frequency domain. Hence each FFT filter must be widened to cover the gaps. This is done by applying time domain window to the incoming data. Also the update rate can also be controlled by selecting proper values of M and N. Polyphase FFT approach allows us to control filter skirts, degree of overlap to meet our system parameters. Polyphase filters are commonly used in Mobile communications for realizing hardware efficient filters for channelization. The present work aims at implementing this approach for ESM application and evaluating it against other proven techniques.

Software used: Matlab.CONTENTS TOPICSPAGE NUMBERAcknowledgement iSynopsisiiChapter 1 HISTORICAL DEVELOPMENTS IN EW1Chapter 2 INTRODUCTION TO ELECTRONIC WARFARE2.1 Introduction62.2 Governing theory13

Chapter 3 DIGITAL RECEIVER SIGNAL PROCESSING SCHEMES3.1 FFT based architecture183.2 FIR filter based architecture183.3 Mixed architecture19

Chapter 4 PROPOSED METHOD20

Chapter 5 METHODOLOGY

5.1 Decimation in the frequency domain21

ConclusionREFERENCES

1.HISTORICAL DEVELOPMENTS IN EW

Electronic warfare is not new; it has been practiced in one from or another in every major conflict since the early days of this century; in fact, ever since radio communications were first used in war. Early techniques were often primitive and it was only from World War II onwards that EW gained an element of sophistication and maturity .Before we go deeper into the study of electronic warfare, let us have a look at the historical development of EW and the strategic role it has played in key conflicts. is will help to highlight the importance of EW.The first reported conflict involving the use of EW was the Russo-Japanese War of 1905, when Russian naval commanders attempted to jam radio transmissions from Japanese ships. However, in this war, the Japanese were successful in trailing the Russian fleet because they could transmit information about their movements and combat formations, without getting jammed, back to the Japanese high command for necessary action.World War I saw the widespread use of radio for communication and transmission of combat information. In 1914, the Germans intercepted the communication system of the British forces. This communication jamming in practice is considered the first real action of EW, as electromagnetic energy had been used, not for communication, but for jamming enemy communications. It was also during World War I that both sides experimented with electronic deception in its the simplest forms, such as false transmissions, electronic the espionage, dummy traffic and other similar ruses for misleading the enemy. Direction-finding achieved great success in maritime operations during this war.However, specialised EW equipment began to be developed only during World War II. Use of radar for war operations was a major development of this period. Early in 1939, Germans employed Luftwaffi's LZI30 Graf up Zeppelin for locating British early warning radars. The Germans also introduced radio guidance techniques for their bombers during night raids on British military installations. Under pressure due to this constant threat of destruction, the British eventually developed a deceptive out. ECM called 'Bromide' against the German technique of eql dropping bombs on pre-located targets. The sophisticated Wurzburg gun-Iaying German radars created a sensation in World War II. The British began to equip their aircraft with both noise jammers and passive ECM equipment as a countermeasure.Throughout the war, there was a fight between ECM and ECCM. Each side momentarily gained the upper hand in EW, only to lose it against a new countermeasure.EW technology became progressively more specialised and sophisticated after World War II. During 'Vietnam War in 1965, Soviet SA-2 'Guideline' radar-guided SAMs (surface-to-air missiles) and 57 mm. radar-controlled AAA (anti-aircraft artillery) made their first appearance in the stormy battlefield, and the first SA-2 downing of a US fighter aircraft was recorded. The US found itself severely short in ECM and early warning equipment to meet this new challenge. To counter this serious threat, some crash programmes were started by the USA to develop an adequate EW capability to reduce the aircraft losses. Like the World War II, the Vietnam War also continued for many years.In 1971, one of history's heaviest barrages of radar-controlled AAA and SAMs were employed against US bombing raids over Hanoi and Haiphong. The US countered by deploying every available EW aircraft including the EA-68 Prowler , he first fully integrated tactical airborne jamming system. Thus the Vietnam War of 1965, which continued up to 1971, clearly demonstrated the conflict between radar and ECM and between ECM and ECCM.The race of developing countermeasures against countermeasures continued to be directed to outmanoeuvre and outperform the adversary's equipment and ECM. The 1973 Middle East War (also known as Yom Kippur War or the Arab-Israel War) saw most of the latest Soviet SAM and AAA systems in action. Each side used a different region of the electromagnetic spectrum for target tracking and guidance. For the first time in modern warfare, a dense ground-based air defence E' system featuring the full spectrum of overlappling SAMs and AAA was thus encountered. In addition to passive communications monitoring, the Arabs employed a range E of ECM and ECCM measures. Poor ELINT (Electronic Intelligence) before the war led to the inadequate preparedness of the Israeli Air Force to counter the Arab air defences. However, aftersustaining heavy air losses in the first few days, they managed to adapt countermeasures to suppress the radar- controlled SAMs and AAA. The 1973 war threw EW into the forefront of modern military thinking. It brought to surface the necessity of possessing a complete range of EW equipment, even in peace time, with an efficiently run Signal Intelligence (SIGINT) service. This war clearly emphasised that if one fails to control electromagnetic B spectrum and to gather intelligence, one may face F disaster.In April 1982, the world saw another EW conflict in the Falklands War. On 4 May 1982,a British Type-42 destroyer, HMS Sheffield, was destroyed by a a sea-skimming French built Exocet missile. The entire military world was shocked, for this type of warship was supposed to constitute the main fleet defence against air attack in cooperation with airborne early warning aircraft to detect low flying enemy aircraft. Unfortunately for Sheffield, there were no airborne early warning radars on board in operation on 4 May when the Exocet missiles were sighted close in. The Sea-Dart missile system on board Sheffield, designed to engage aerial platforms at a distance, was simply unable to get on target in time because of its slower reaction time. Assessing the war from the point of view of EW, several innovations were employed in ground combat. Skeffield lacked the latest EW equipment capable of countering technologically advanced western missiles like the Exocet. Argentinian forces made very little use of EW systems, except passive EW, like ELINT and ESM. Excellent organisation of command, control, communications and intelligence (C31) on the part of British forces contributed significantly to its success in the Falklands War.In June 1982, another fierce EW battle, known as the Lebanon War, was fought between Israel and Lebanon in the Bekka Valley . By mid-june, Israeli forces reported the destruction of 86 Syrian aircraft, including Soviet-built 'Makoyan' MIG-23 fighters and five French-built ' Aero Spatiale Gazelle' attack helicopters. The Israelis reported that they, in turn, had lost only two helicopters and the Bekka Valley area had been destroyed by the Israeli Air Force without much losses. In this war, the Israelis made use of a special type of deception technique, called decoys (drones and RPVs, remotely piloted vehicles) to know the location and characteristics of enemy operating systems and weapons. Israelis had employed their 'Mastiff RPV (as well as drones) to ascertain the microwave radio frequencies used by the Syrian SAM- 65. Two Israeli Grumman, E-2C Hawkeye aircraft obtained electronic bearings of the Syrian missile radar system, allowing them to plot their exact location. Israeli aircraft then destroyed the sites with rockets riding a microwave beam to the SAM-6 sites. This led to a stunning defeat of Lebanese forces and an incredible victory of Israeli forces. Israeli Air Force played a dominant and decisive role In this war. What was the decisive factor? Electronic warfare.The outstanding results achieved by the Israelis show that the new concept of real-timewarfare, supported by accurate planning of EW actions, was the real key to their success. Another element which contributed greatly to the Israeli success in Lebanon was the coordinated use of AWACS (Airborne Early Warning and Control System) and ECM against enemy command, control and communications systems, called C3CM. The classic struggle between the lance and shield, the gun and armour, the missile and electronic systems, the countermeasures and counter-countermeasures will no doubt continue in the form of fight between radiation weapons and radiation countermeasures and between these counteremeasures and relative counter-countermeasures and so on.Electronic warfare today is an utterly deadly battlefield, where victory or defeat may come in a matter of seconds, even microseconds. Thus, in this situation inadequate EW means certain defeat.Many technological developments have come about as a result of military needs. An engineer was originally a person who designed and built military fortifications and equipment. But during the World War II, the intellectual forces of scientific research and development were deliberately and intensively applied to the conduct of war. Winston S Churchill was one of the first political leaders, with no significant background in science, to have engineers and scientists as advisors, to listen to them and utilise his political power to translate their scientific knowledge in to practical wartime technology .He was also the first leader to recognise EW as a vital phase of military operations.Since the end of World War II, EW has been one of the best kept secrets with technicalexperts and armed forces. It is still in the interest of these two groups of people, though for different reasons, to keep EW developments hidden from indiscreet and unscrupulous people. For a crew of military aircraft, a tank or a warship, an appropriate EW tactic which has been kept secret can mean the difference between the success and failure of their mission, or even the difference between life and death. Therefore, there are strong reasons for keeping many aspects of EW secret. However, there are equally strong reasons for not only the armed forces and those concerned with National Defence but also the academicians, students and the general public being informed about the existence and general usefulness of EW. Upto World War II, Radar was the most secret weapon in the hands of military forces. The common man was interested to know about its capabilities. Soon after World War II all secrets of radar were thrown open to the public. It soon became a household word. It revolutionised the areas of research in the academic and specialised research institutions. Today, we have a great variety of radars performing thousands of functions for war, peace and public good.2.INTRODUCTION TO ELECTRONIC WARFARE

2.1. Introduction

An EW system is used to protect military resources from enemy threats. The field of EW is recognized as having three components: Electronic support measure (ESM), which collects information on the electronic environment. Electronic countermeasures (ECM), which jam or disturb enemy systems. Electronic counter-countermeasures (ECCM), which protect equipment against ECM.Because it does not radiate electromagnetic (EM) energy, the first is often referred to as a passive EW system. The second is referred to as an active EW system, since it radiates EM energy. Because they do not emit EM energy, such techniques as stealth targets (that avoid being detected by enemy radars) and deployment of decoys or chaff (thin metallic wires) to confuse enemy radars are also considered as passive EW. ECCM is usually included in radar designs; hence, it will not be discussed here.EW intercept systems can be divided into the following five categories:1. Acoustic detection systems are used to detect enemy sonar and noise generated by ship movement. These systems detect acoustic signals and usually operate at frequencies below 30 kHz.2. Communication intercept receivers are used to detect enemy communication signals. These systems usually operate below 2 GHz, although a higher operating frequency is required to intercept satellite communication. These receivers are designed to receive communication signals.3. Radar intercept receivers are used to detect enemy radar signals. These systems usually operate in the range of 2 GHz to 18 GHz. However, some researchers intend to cover the entire 2 to 100 GHz range. These receivers are designed to receive pulsed signals.4. Infrared intercept receivers are used to detect the plume of an attacking missile. These systems operate at near through far infrared (wavelengths from 3 to 15 /mm).5. Laser intercept receivers are used to detect laser signals, which are used to guide weapon systems (i.e., attack missiles).The intercept receivers often operate with EW signal processors. The processors are used to process the information intercepted by the receivers to sort and identify enemy threats. After the threats are identified, the information is passed to an ECM system. The ECM system must determine the most effective way to disturb the enemy operation, which may include throwing out chaff. The actions of an ECM system against radars include noise and deceptive jamming. Noise jamming is intended to mask by noise the radar's return signals from targets so that the radar cannot detect any signal and its screen is covered with noise. Deceptive jamming creates false targets on the radar screen such that the radar will lose the true targets.

The architecture explored in this research specifically focuses on a wideband digital intercept receiver that is capable of exploiting both radar and communications signals. More emphasis is placed on detection of radar, however, since radar signalsgenerally have wider bandwidths and operate in a larger portion of the EM spectrum.Communication receivers, on the other hand, historically use narrower bandwidths.Nonetheless, since the basic digital hardware components that comprise radar (wideband) and communication (narrowband) intercept receivers are similar, it is important to distinguish the differences between these types of systems.

2.1.2 Wideband EW vs. Narrowband Communication Receivers.

Wideband and narrowband digital receivers are two important types of systems used in EW and communications. Historically, wideband digital receivers have been used to intercept wideband pulsed radar signals, and narrowband receivers have been used to receive communication signals. Due to advances in digital signal processing techniques, the basic components of these types of receivers are very similar, and for a digital receiver design as discussed in this research, the architecture could be applied to either type of system.Wideband receivers have a very wide instantaneous input bandwidth of about 1 GHz or larger. This allows any signal in the bandwidth of interest to be intercepted instantaneously without needing to tune the receiver. This is important for detection of pulsed radar signals emitting Linear Frequency Modulated (LFM) waveforms capable of more than 200 MHz bandwidth. Wideband waveforms emitted by radar is important since the wider the bandwidth, the more accurate the range measurement to the target. This is crucial for high resolution imaging used in synthetic aperture radar (SAR). Unfortunately, the larger the bandwidth, the shorter the width of the pulse, so the wideband digital receiver must also be capable of sampling the spectrum rapidly to ensure no pulses are missed and the Time of Arrival (TOA) of a pulse is accurately calculated.

Another advantage of wideband receivers is that having a wide instantaneousbandwidth can allow tracking of multiple signals simultaneously. Subsequently, wideband digital receivers are also used for analyzing the EM spectrum for any signals of interest. Unfortunately, wide bandwidth can mean less dynamic range since there is more noise in the spectrum due to the larger bandwidth . Dynamic range is an important factor in the ability for the receiver to characterize signals in the presence of noise and spurious frequencies. As a result, a wideband receiver can be used to provide coarse signal information over a large instantaneous bandwidth and direct a narrow band receiver, with a larger dynamic range, to obtain a fine measurement on the input signals. This is the primary function of a wideband queuing receiver . Narrowband receivers, commonly have a much narrower bandwidth, and are mainly used as communication receivers .Examples of narrowband communication receivers include, AM, FM, and television channels which occupy 10 kHz, 200 kHz, and 6 MHz bandwidths respectively. With the recent advent of wireless broadband communications the instantaneous bandwidth of communication receivers is approaching what is considered wideband . Operating modes such as Ultra-Wideband (UWB), multi-band Orthogonal Frequency Division Multiplexing (OFDM), and spread spectrum are found in consumer wireless products which use these wideband techniques to increase communications throughput .In fact, UWB bandwidths of 1.5 GHz and greater are being developed in consumer products for indoor use .Despite the fact that narrowband communication receivers are approaching wideinstantaneous bandwidths, there are two important aspects that distinguish them from wideband EW receivers. First, in the design of communication receivers, the frequency, modulation, and bandwidth of the incoming signal is known. As a result, the communication receiver can be designed optimally for the input signal .Even a radar receiver can be considered a communication receiver because the received signal is known to be a function of the transmitted signal and a matched filter is used on the radar receiver to maximize detectability of a target. Wideband EW intercept receivers, on the other hand, must operate in an environment where the information of the input signal is unknown. In addition, a transmitter may use spread spectrum techniques such as pulsed FM chirp, polyphase coded signals, and frequency hopping to avoid detection by an intercept receiver . Nonetheless, these signals are generally much simpler to detect than communication type signals .Another major difference between narrowband communication receivers andwideband EW intercept receivers is EW receivers output pulse descriptor words (PDWs), which describe the characteristics of a detected signal. Such characteristics include frequency, angle of arrival, pulse width, pulse amplitude, and time of arrival on each received pulse . Narrowband communications receivers, on the other hand, recover information emitted by the transmitter, such as video for TV, or audio for FM. In addition, a majority of communication receivers receive continuous wave (CW) signals, whereas EW receivers are designed primarily to receive pulsed radar signals (but can still exploit CW signals). For this reason, TOA is a much more important metric for wideband EW receivers than it is for communication receivers. Consequently, the focus of a wideband EW receiver is the fast sampling of the EW spectrum for signal detection and characterization, whereas for the communication receiver, it is simply the recovery of information from a specific type of signal.

2.1.3 The Analog Wideband EW Receiver.

The concept of a digital wideband EW receiver is generally recent. Conventional EW receivers are primarily made up of analog components. In addition, there are many different architectures for EW receivers based on desired requirements, such as sensitivity, input signal range, dynamic range, response time, and how many simultaneous signals can be detected .With recent advancements in ADCs, recent research has concentrated on using digital receiver architecture to replace many of the traditional analog functions of conventional receiver design.A basic diagram of a conventional analog EW receiver architecture is shown in Figure 1. The EW receiver itself consists of an RF section and a parameter encoder. The antenna at the left captures pulsed RF signals that are generated by most radars. RF ranges from 2 GHz to 100 GHz, but for radar, the most popular frequency range is from 2 to 18 GHz. Next an RF converter down-converts a high frequency signal into a lower intermediate frequency (IF) signal so that the EW receiver can more easily process the same bandwidth signal at a lower frequency. The signal then proliferates to the RF section that can contain further signal conditioning devices such as filters and amplifiers. In most analog EW receivers, the RF section contains a diode envelope video detector which converts the RF signals into video, or DC signals . Once the signal is converted to a video signal, it proceeds to the para (or parameter) encoder.

fig: 1: the conventional EW receiver

The para encoder is responsible for outputting a digital word describing the parameters of the signal. This is also known as a pulse descriptor word. The parameters of a PDW can contain pulse amplitude (PA), pulse width (PW), time of arrival (TOA), carrier or RF frequency, and angle of arrival (AOA) . Not all EW receivers can output all five parameters. Generally, receiver design problems occur in the parameter encoder design, which is subject to deficiencies, such as reporting erroneous frequency [29]. As a result, a satisfactory encoder is difficult to achieve.The digital processor is responsible for gathering PDWs from the parameter encoder. Further processing is done on these PDWs to fully characterize signals of interest.

2.1.4 The Digital Wideband EW Receiver.Recent research in general receiver design has focused on using digital signal processing to replace many of the functions of analog components. With advancements in ADCs and the increase in digital processing speed, digital receivers are the primary types of receivers being researched and designed today .The primary reason for trading analog functionality for digital domain processing is that digital signal processing (DSP) has performance advantages related to manufacturability, to insensitivity to the environment, and a greater ability to absorb design changes compared to analog counterparts . High quality analog components have to be manufactured with tight tolerances and are always performance dependent on environmental factors such as temperature. As a result, there is a high impact on cost. Digital components can be designed or even reconfigured to work with a myriad of requirements using a generic hardware architecture. Such is the idea of software defined radio as discussed in . Nonetheless, analog components are still a necessity for high frequency RF signal processing since ADCs that can directly sample the 2-18 GHz band and higher are scarce and are primarily in the research stage.A basic diagram of a digital EW receiver is shown in Figure 2. Compared tothe analog EW receiver, the video detector is replaced with an ADC, which outputsdigitized data that are samples of the incoming signal in the time domain . This data must then be converted to the frequency domain with some type of spectral estimator. Spectral estimation is commonly done using a digital Fast Fourier Transform. The digital representation of the frequency spectrum is then output to the parameter encoder which generates PDWs containing the five parameters.

fig: 2 : the digital EW receiver

2.1.5 Importance of TOA for EW receivers.

Time of Arrival is a PDW parameter that assigns a time tag to the leading edge of a received pulse at the receiver input . The TOA information is used to determine the pulse repetition frequency (PRF) of a radar. In addition, TOA is used to deinterleave multiple pulses having different pulse widths that arrive at the receiver from multiple radars. The typical duration of radar pulses may be anywhere between tens of nanoseconds to hundreds of microseconds, and the PRF can range from a few hundred hertz to about one megahertz. TOA accuracy is directly dependent on how quickly the EW receiver can sample the EM spectrum. For radars with high PRFs and/or small pulse widths, it is important for the receiver to be fast enough to detect the rising edge of the pulse. Otherwise, the receiver may only detect half the pulse or miss it entirely if the spectrum sampling rate is too slow. A simplified example of a transmitted radar waveform is shown in Figure 3. Each pulse has a pulse width of PW , and contains a sine wave with a carrier frequency fc. In addition, each pulse repeats at a specified interval of time known as the Pulse Repetition Interval (PRI). The frequency at which the pulse repeats is the PRF, where PRF = 1/PRI. In modern radar, rather than a simple sine wave, more complex waveforms like LFM, polyphase, and binary phase-coded waveforms are used to increase the bandwidth of a pulse . Consequently, wideband, high update rate EW receivers are definitely needed as radars are designed to transmit shorter pulses with more complex wideband waveforms. For a digital EW receiver, the spectrum estimator (normally an FFT) is usually the limiting factor in update rate due to its computational complexity. Normally, a set number of samples is captured and processed to produce a spectrum. For an FFT, the number of samples captured is directly related to how well the signal can be resolved in frequency. The frequency resolution is defined as Fs/N, where Fs is the sampling frequency of the ADC, and N is the number of samples taken. The spectrum update rate is simply the reciprocal of the frequency resolution, N/Fs. For example, the Monobit receiver, as described in , processes 256 samples of ADC data at a sampling rate of 2.56 Giga-samples/second (Gs/s). This yields a frequency resolution of 10 MHz, with a spectrum update period of 100 ns. In this case the TOA resolution is 100 ns, and the minimum desired pulse width is also 100 ns.

fig 3: basic radar pulse

2.2 Governing TheoryA basic knowledge of digital signal processing is needed to understand the theory behind the channelized wideband digital receiver design presented in this research.The architecture presented includes aspects of the discrete Fourier transform, filter design, windowing, and multirate system design.

2.2.1 The Discrete Fourier Transform. The discrete Fourier transform (DFT) is a way to represent a sequence in terms of a linear combination of complex exponentials for a finite-length sequence [9]. The DFT corresponds to a sequence of samples that are equally spaced in frequency of a Fourier transform of the signal.This is important in the digital domain, since it is a one-to-one mapping of a time sequence x(n) to another sequence X(k) representing complex frequency. The mapping is shown as:

The DFT is rarely implemented as shown due to its computational complexity. The DFT equation has a complexity of O(N2), which, for an N-point DFT would need N2 complex multiplications and N (N 1) additions . This would require an inordinate amount of hardware resources for a reasonable DFT of size N. For this reason, the fast Fourier transform is commonly used instead.

2.2.3 FIR Filter Design by Windowing. Filtering is an important and necessary operation in any wideband EW receiver. A filter is a system that ultimately alters the spectral content of input signals in a certain way. Common objectives for filtering include improving signal quality, extracting signal information, and separating signal components . Filtering can be performed in the analog or digital domains. Since the focus of this research is on a digital receiver, a digital filter design is explored.The filter designed in this research is a finite impulse response (FIR) filter. FIR filters are commonly used in DSP implementations for a variety of reasons. Most importantly, FIR filters are linear phase filters, so phase distortion is avoided . Inapplication to a wideband EW receiver, this is important since many times, the AOA calculation is reliant on the phase difference between multiple channels. Use of FIR filters are also desirable because they are always guaranteed to be stable due to their absence of poles in the transfer function. FIR filters are feed forward filters, and do not utilize feedback like infinite impulse response (IIR) filters, which can produce instability especially when considering coefficient quantization errors . FIR filter design is primarily performed using the windowing method. The windowing method is commonly used for spectrum analysis as well as filter design.

2.2.4 Multirate Systems and Filter Banks. Multirate signal processing is a large subject of study that is very important for implementation in systems for speech analysis, bandwidth compression, communication, and radar and sonar processing .Multirate techniques are used to primarily increase the efficiency of signal-processing systems, which is a very important consideration when implementing in hardware .For this research, polyphase filtering and decimated DFT filter banks are implemented in this design to increase the spectrum update rate, thereby reducing TOA for digital EW receivers.The fundamental idea of multirate systems is a change in the sampling rate of a system. Unlike a single-rate system, a multirate system allows sampling rates to be kept as small as possible at certain points throughout the processing sequence, therefore, yielding more efficient processing . The two basic components of a multirate system are the downsampler (or decimator) and the upsampler (or expander). The downsampler takes input samples at a high rate and produces a reduced output rate by an integer factor. Conversely, the upsampler takes as input a slow data rate and increases the effective sampling rate.

fig:4: downsampler and upsampler

For the purpose of this research, only the downsampler is utilized to reduce the internal data rate of the digital receiver as it processes the incoming ADC data. The mathematical representation of downsampling is :y(n) = x(Mn)where M is an integer, and only those samples of x(n) equal to multiples of M are retained.

Polyphase filtering is a multirate signal processing operation that leads to an efficient filtering structure for hardware implementation.Polyphase filters are also used to sub-channelize data for filter banks. Filter banks are used for both communications and spectral analysis as applied to EW receivers. Examples of two different filter banks is shown in Figure 5. The overlapped filter bank (a) is useful for applications that require no missed frequencies. In this case, a narrow-band signal can straddle one or more output channels, but there is no risk of losing the frequency [8]. The nonoverlapped (b) filter bank is useful in communications, where separation of the output channels is crucial in mitigating cross-channel interference.

fig 5 : Filter Banks with a) overlapping and b) nonoverlapping bands

The specific filter bank explored in this research is the polyphase DFT filter bank. A diagram showing the polyphase DFT structure is shown in Figure 6.

fig 6: Polyphase DFT Filter Bank Structure

3. DIGITAL RECEIVER SIGNAL PROCESSING SCHEMES

3.1 FFT-based ArchitecturesFFT-based architectures use the Fast Fourier transform algorithm for signal filtering and signal detection. A given N point FFT divides the input frequency band into N fs/N filters. The shape of the filter is decided by the time domain window used. Parameter measurement is done on the FFT outputs. The FFT outputs are complex. Hence pulse envelope can be calculated using the complex magnitudes and instantaneous frequency can be calculated using differential phase measurement. FFT based architectures can generally be pipelined and can handle very high pulse density. Selection of number of points of FFT determines the frequency resolution and the TOA resolution. Also for accurate PW measurement through interpolation, it is required that time frames to be overlapped for finer TOA estimation without degrading the frequency resolution. But this requires extra storage and multipliers than the non-overlapped FFT case.

3.1.1 Non-overlapped FFT ArchitecturesIn this scheme, the input data is divided into frames of N samples before performing N-point FFT. On the output of consecutive FFTs, the signal detection is performed.

3.1.2 Overlapped FFT ArchitectureIn this scheme, the input data is overlapped before performing the FFT. We consider the case of data overlap of 128 and 192 points. The data overlap increases the FFT output rate and hence increases time resolution without degradation in frequency resolution. The implementation complexity grows linearly with amount of overlap.

3.2 FIR filter based architectures3.2.1 Cascaded FIR filter based architectureCascaded FIR filter architecture use bank of FIR filters in stages to achieve desired frequency resolution.The given input band is divided into M wide filters. Each M filter is further sub-divided into N narrow filters. The input signal is multiplied with the center frequency of each filter to bring it to the center of the filter. Threshold is applied to each filter output for detection.

3.3 Mixed Architectures3.3.1 FFT + FIR Filter ArchitectureIn this scheme, the pulse detection is based on 256 point non overlapped FFT. Based on the detection in the FFT, Up to Four FIR filters can be tuned based on four maximum peaks in the FFT. The parameter measurement is done using 4 I and 4 Q decimator filters. Prototype I and Q filters are designed and stored. Based on the FFT detected frequency, I and Q decimator filters are tuned to the frequency of the signal and signal samples are filtered as well as decimated. This scheme requires external or internal storage memory for storing raw ADC samples. All the pulse measurements are done on filter outputs. This scheme gives better time domain resolution than asimple FFT as a filter operates sample by sample.

4. PROPOSED METHOD

4.1 Polyphase FFTPolyphase filtering is a multirate signal processing operation that leads to an efficient filtering structure for hardware implementation. Polyphase filtering parallelizes the filtering operation through decimation of the filter coefficients, h(n). This allows a lower internal processin rate with shorter filters that yields larger effective rate of execution. Polyphase filters can also be used to sub-band the frequency spectrum, thus producing a filter bank. Figure 7 shows the implementation of Polyphase FFT.

fig 7: Polyphase FFT architecture

Design Procedure : The incoming N data samples are distributed in M branches and an M point FFT is performed. Number of points M is selected based on time resolution required. The decimation results in gaps in the frequency domain. Hence each FFT filter must be widened to cover the gaps. This is done by applying time domain window to the incoming data.

5. METHODOLOGYThe wideband digital receiver presented in this research is designed using a uniqueapproach of decimation in frequency to allow a trade between frequency resolutionand update rate to improve the time of arrival estimate. By designing thearchitecture generically and defining a set of design parameters, system level engineerscan generate wideband digital receiver architectures to suit the specific needs of theEW system.Figure 8 shows a simple block diagram of the Channelized Wideband Digital Receiver. The digital receiver consists of a decimation filter, FFT, and encoder/signal processor that outputs a PDW. Ideally, these components would reside in a FPGA or ASIC. The shaded block contains the decimation filter and FFT. The shaded block is realized as a polyphase DFT which uses decimation in frequency to filter the incoming ADC input data and produce a frequency spectrum as output. In this chapter, the theory of decimation in frequency is discussed as well as the design flow used to go from mathematical simulation to hardware implementation.

fig 8: Channelized Wideband Digital Receiver

5.1 Decimation in the Frequency DomainDecimation in the frequency domain is a processing method in which the frequency spectrum is decimated. Normally, the decimation is performed as a time domain operation in which input samples are thrown away by an integer factor to reduce the output rate. Time domain decimation has the frequency domain effect of reducing the output bandwidth. Since the output bandwidth is reduced, filtering must be performed to avoid aliasing since the effective nyquist rate changes from Fs/2 to Fs/2M where Fs is the sampling rate and M is the decimation factor. For decimation in the frequency domain, the frequency values are decimated which has an effect of reducing the complexity of the FFT operation.

5.1.1 Mathematical Description. For this description, a specific case of frequency domain decimation is presented since the notation is simplified and the concept is straight forward. The general set of equations are defined later. Most of the following derivation is found in. For the specific case, we assume a 256-point FFT with a decimation factor of 8. The DFT equation is written as:

If N = 256, there are 256 outputs in the frequency domain, so with a decimation factor of 8, every 8th output is kept. Consequently, the results for k = 0, 8, 16, ..., 248 are calculated and there are a total of 32 (256/8) outputs. These outputs can be written as:

Next, X(k) for each value of k is written in a slightly different form. An example for X(16) is:

In this form, it can be seen that the DFT operation for one frequency component can be split up arbitrarily into groups of data points that are added together and multiplied by complex exponentials. For this case, the operation is organized into 32 groups of 8 data points, each multiplied by a different exponential as determined by k and n. If the decimation factor were 4 instead of 8, there would be 64 groups of 4 data points instead. The bracketed values in above equation can be defined as a newquantity, y(n), as:

where n = 0 to 31. Therefore, each y(n) value contains 8 data points. Using y(n),the FFT outputs :

These equations can then be described by:

where k = 0, 1, 2, ..., 31 and n = 0, 1, 2, ..., 31. X(8k) represents the decimation of theoriginal frequency spectrum by a factor of 8 and is relabeled as Y (k). Equation now describes a 32-point FFT of the time sequence y(n).For the generalized case, if an N-point FFT is desired, and the outputs of frequency spectrum are chosen to be decimated by a factor M, only an N/M-point FFT is required. In order to accomplish the frequency domain decimation, the input data, x(n) must be reordered and processed to produce y(n) as input to the N/Mpoint FFT. The generalized equation for y(n) is:

where n = 0, 1, 2, ..., (N/M)1. Once y(n) is calculated, the outputs in the frequencydomain are obtained by:

Above equations summarize the process required to perform decimation in the frequency domain. The apparent advantage is the ability to determine the spectral output of an N-point FFT while only requiring an N/M-point FFT, resulting in a significant computational reduction. Though there is a computational advantage to performing decimation in frequency, the tradeoff is a reduction in frequency resolution.

5.1.2 Channelization through Polyphase FilteringThe channelized polyphase filtering method has some significant advantages.First, by parallelizing the filter through polyphase decomposition, the sampling rate of each individual filter is reduced by a factor of 1/D, where D is the number of filters.

fig 9 : 256 pt. Channelized Polyphase FilterFor our special case, if the sampling rate were 2000 MHz, the polyphase filters would only need to run at 2000/32 = 62.5 MHz. If we were to use a 256-tap filter for windowing, the filter would need to run at the full sampling rate. This fact is very important when considering implementation of the filter in hardware. Generally, as the size of the filter increases, the maximum sampling rate at which it can run in hardware decreases . For this reason, polyphase filtering yields a large performance increase for hardware implementation since it can perform the same operation at a lower sampling rate. Consequently, the polyphase filter is capable of processing more data at the same clock rate as a normal filter.A second significant advantage to using the channelized polyphase filtering method is an increase in time resolution, which improves the TOA and PW calculations in an EW receiver. For our example, the polyphase filter and FFT processes data 32 samples at a time. This yields a time resolution of 32/2000 MHz = 16 ns. A normal 256-point FFT must wait for 256 samples before it can process a spectrum, which results in a time resolution of 256/2000 MHz = 128ns. As a result, by decimating by 8 and using polyphase filtering method, an 8-fold performance increase in time resolution is gained.

It is important to note that frequency resolution and time resolution have an inverse relationship described by Tres = 1/Fres, where Tres is in seconds and Fres is in hertz. By decimating the frequency spectrum, the frequency resolution decreases and the time resolution increases according to the decimation factor. For our example, when the frequency spectrum is decimated by 8, the frequency resolution and time resolution change from 7.8 MHz and 128 ns to 62.5 MHz and 16 ns respectively.

CONCLUSION

polyphase FFT can be a good processing scheme for future digital receivers. Future work in this area could be exploring IIR filters for their suitability for digital receiver operations

REFERENCES Approaches towards implementation of multi-bit digital receiver using FFT, Abhijit S Kulkarni, Vijesh P, Hemant V Paranjape, K Maheswara Reddy. Defence Science Journal, Vol 63, No.2 , March 2013,pp.198-203, DESIDOC. Paramaterizable channelized wideband digital receiver for high update rate, Buxa Peter, Wright state university, 2003(MS Thesis). DARE Digital Receiver Design Document VER 1.0, DARE 2010. Electronic Warfare, Mohinder Singh, Institute of Armament technology, Pune,DESIDOC. Digital techniques for Wideband digital receivers Ed 2, Tsui James B Y, Norwood, MA: Artech house, 2001. Lyons, Richard. Understanding digital signal processing. Pearson Education, 2004.