With the advent of ever smarter and faster anti-ship missiles, many navies around the world have shifted their focus to the underwater dimension for credible offensive maritime capability. In the Indian Ocean Region (IOR), the Indian Navy (IN) has therefore decided to focus heavily on staying at the cutting edge of anti-submarine warfare (ASW) capability especially in light of increasing sonar contacts with Chinese attack submarines both diesel-electric (SSK) and nuclear (SSN). The cutting edge element of successful ASW operations however comes from potent intelligence surveillance and reconnaissance (ISR) technology relevant to submarine tracking and detection. Modern underwater detection systems obviously leverage digital technologies to both increase the efficacy of sensors by improving the signal to noise ratio (SNR) as well as to facilitate network-centric operations to using dispersed sensor nodes.
The digitisation of sonar systems over the years has allowed them to stay in the game despite appreciable advances in radiated noise management for submarines. Indeed it could be argued that passive sonar would have lost its relevance had it not been for the arrival of digital signal processing (DSP) and array beam forming (ABF) techniques on the scene. Passive sonar essentially involves processing the sound signal generated by the target for estimating its bearing and characteristics through spectrum analysis. Now received noises with SNR of -15dB or below are very common at the receiver for passive sonars. When one adds reverberation and scattering, especially in modern littoral environments plus the usual issues related to variable sound speed due to salinity and refraction on account of thermoclines, it is evident that without high-end data processing passive sonar would have struggled to stay contemporary with threats despite the fact that passive sonar does not betray its own location to a potential contact.
For most of the 20th century highly trained sonar officers classified and recognized targets by actually listening to their radiated noise. Given the complexity of today's environment this has been substituted by microprocessors running DSP schemes increasingly based on hybrid methods such as a combination of hidden markov methods and artificial neural networks. This has considerably improved the detection and classification of targets by passive sonars. The 'gain' required for target tracking in passive sonars is of course provided through the use of beamforming techniques.
Naturally all this has greatly increased the need for computational power coupled with huge memory and large I/O bandwidth to run DSP methods on the information extracted which includes direction of arrival, speed of the contact, the bearing rate, dominant tonal frequencies, shaft rpm, and the number of blades. As such high performance digital signal processors and PowerPC-based boards are now standard hardware found in the sonar back-ends. For instance in sonars developed by DRDO's Naval Physical & Oceanographic Laboratory (NPOL), Kochi, DSP functions have been implemented using high-speed digital signal processor boards based on open standards using TMS processors.
Processor speed is of course upgraded over time. The latest TMS for instance offers 160 GFLOPs of peak performance per device with eight cores and is found in pairs on a 6U VPX form factor board. This means that the total peak processing power of the board is 320 GFLOPs and it supports 1 GB of DDR3 memory and 128 MB of NAND Flash and 32 MB of NOR Flash. The total power dissipated by the board is approximately 80 W. Just a few years ago the maximum speed of such a board was only about 24 GFLOPs thereby indicating that while the performance of the overall electronics it rising, the space it occupies is getting downsized through what is being called 'hardware compression'.
However even as the processing back end leverages the joys of the digital revolution, advances in reception hardware have also made a difference. A sonar is ultimately as good as the transducers that convert acoustic signals to electric ones for the signal conditioning unit. In the case of a passive sonar, acoustic transducers today are broadband omni-directional hydrophones that merely listen to underwater sounds and are mostly made of piezoelectric ceramics, but could also be made of magneto-restrictive or electro-dynamic materials. For deep water hydrophones i.e for use on submarines themselves, acceleration balanced technology is considered mature with deep water passive arrays being able to operate at depths of up to 600 metres and at a frequency of 10 Hz.
Now transducers can not only be hydrophones i.e receivers but be projectors for underwater transmission as well. Transducers of the projector variety are of course the relevant type for active sonars that emit pulses of sound waves that travel through the water and process the received target echo to estimate the range, bearing, and Doppler of the target.
The rise of modern quietening technologies started a trend a decade and a half ago whereby navies today mostly field low frequency (LF) active bow and hull mounted sonars operating in the 100 Hz to 1 kHz bands, with passive sonars now mostly found in towed array, dunking sonar and sonobuoy configurations. The range performance of LF active sonars is far superior to those of passive ones as they are less affected by propagation mechanisms that scatter high frequency signals. Moreover active sonars are also sought for their ability to defeat thin anechoic coatings on target submarines.
However submarine stealth is not a one-time game and given the need to detect them beyond the range of their heavyweight torpedoes has meant that active sonar architecture too is moving towards towed arrays and dunking sonars. The winch and cable configuration keeps a towed array's sensors at a distance from the ship's own noise sources, greatly improving SNR, and thereby increasing potency in detecting and tracking quiet, low noise-emitting submarine threats, or seismic signals.
In terms of transducer technology this is leading to the adoption of LF flextensional type transducers instead of the tonpilz type since the former has much better power handling capacity, and power-to-weight and power-to-size ratios. Either type can of course serve as both projector or hydrophone. The transducers in hydrophone mode obviously receive the returning signal from the target. In fact most active sonars today also have passive modes (active cum passive) both in hull mounted types such as NPOL's HUMSA-NG or in towed array types such as NPOL's ALTAS which is heading for final user evaluation trials in 2015. In the future as thin line towed arrays for small unmanned sea vehicles (USVs) become common, the focus will shift to micro-electromechanical systems (MEMS) based transducers that are miniaturized sensors integrated with signal conditioning, interface circuits and other electronics.
Active sonar technology is of course also utilizing the DSP technology and processor technology mentioned above. However given space constraints power amplifier (PA) technology too is evolving for active sonars with a focus on compactness and energy-efficiency which is why new sonars are mostly using IGBT or MOSFET based PAs.
The overall submarine detection game is however also moving towards increasingly networked arrays in the form of sensors in multi-static mode whose data can be shared with each other or fused together at a centralized decision junction. Networking facilitated by the ability of individual sensors to channelize ever increasing amounts of data allows signals to be integrated over a wide area which can be harnessed by both local and/or centralized processing. Many constituents of such a network could easily be USVs some of which may operate passively while others may serve as a noise source. After all passivity does have the advantage of not giving away a sonar's location to a target.
Wide area ASW sensing therefore has the potential to render choke points and littorals far more difficult for submarine operations. Especially if combined with sonobuoys carrying GPS locaters which are typically deployed from airborne vectors. The data from the GPS module is collected using a microcontroller and is transmitted through antenna in frequency shift keying (FSK) format. The receiver decodes FSK input and sends it through an ethernet interface. With GPS data for every sonobuoy now available beamforming techniques using tomographic algorithms under development turn the entire sonobuoy array into a large steerable virtual hydrophone leading to more accurate bearing measurements and increased sensitivity.
Any network however is only as good as its wireless communications system. NPOL's 3-G Underwater Wireless Acoustic Communication System (3G UWACS) now upgraded to a product called Triton, is one such example. Triton, according NPOL offers a tunable wideband communication capability using a software defined radio (SDR) architecture over multiple bands in voice and data communication modes between surface units and submarines. This system incorporates advanced modulation and coding techniques in addition to data recording and analysis features. It offers the user flexibility in operation and supports remote operation and monitoring through standard networking technologies critical for platform-level integration. The system can be utilized in stand-alone or integrated modes of operation.
It is clear that heading to the future underwater detection systems are going to become more networked with several small unmanned platforms operating in concert with defending ships and submarines to develop a common underwater picture for operations. This dispersal of assets will make sneaking in even small and quiet SSKs into choke points and littorals very difficult. It is no wonder that those navies that can afford to are looking to increase SSN holding with an emphasis on long range sensing and firepower.
For updates follow Saurav Jha on twitter @SJha1618.
