A smart sound receiver device as part of a network-centric system for establishing the coordinates of a sound source

Fomichev V.A., reserve colonel,
Kryukov V.A., civilian.

Alive in the help of the Most High,
will dwell in the shelter of the Heavenly God…

Psalm 90.

This article is a continuation of the article “CAN bus based smart sound sensor receiver design concepts»

The authors of this article have joined forces to develop a compact, low-resource, low-cost network-centric counter-battery system to protect people's lives.

The article presents our understanding of the design of smart sound receivers and their joint work as part of a network-centric system to establish 3-dimensional coordinates of the sound source.

The use of smart sensors is also possible in civilian areas.

1 Problem status

The basis of existing methods for determining the coordinates of the battery of the opposite side is the method proposed by Lieutenant of the Preobrazhensky Regiment N.A. Benoit in 1909, called the time difference method. It is described in detail in an open source [1].

In sound reconnaissance, the determination of the coordinates of targets (benchmarks) is carried out by approximately finding the bearing to the target by solving a sound metric formula based on data from at least two acoustic bases. The acoustic base consists of at least two sound receivers located at some distance from one another. Bearings are determined by the difference in time (report) of arrival of the wave front at the sound receivers.

For more than a hundred years, the acoustic method of determining the coordinates of a point sound source has virtually not undergone significant improvements. Its significant limitation is:

· binding of the desired point to hyperbolic asymptotes, which leads to an obvious error in the location of the sound source and leads to a trail of specific correction methods.

· initial two-dimensionality of the method.

A detailed overview of modern special technical means for determining the coordinates of a sound source is given in [2].

2 Element-by-element description of the sound receiver

The block diagram of a smart sound receiver is shown in Fig. 1

2.1 Glonass receiver with remote antenna

receives signals from Glonass global positioning satellites (GPS, Galileo) and determines:

· geographic coordinates of the receiver,

altitude above sea level and

· exact local time.

Data is transmitted via the CAN bus to the Fast Fourier Transform (FFT) block in NMEA text format via the UART interface.

Although the accuracy of the time obtained from Glonass is on the order of tens of ns, the UART speed may not be sufficient to generate an operational time stamp. An error in determining the time of arrival of a sound wave of 1 ms gives a component of the error in calculating the coordinates of the sound source by 1 order. It is necessary to ensure the accuracy of recording the moment of arrival of the sound wave front at the FFT block to several microseconds. Therefore, the use of time received from Glonass is limited to the synchronization of the internal clocks of the receivers.

When placing the receiver mobile on a vehicle, the UART speed should also be taken into account to accurately determine the coordinates of the moving receiver.

Glonass receivers are presented in a wide range of navigation devices on the market. Also available with a CAN bus adapter.

2.2 FFT block with microphone

Fast Fourier Transform Block

Converts the audio analog signal into digital samples,

· processes the FFT digital data stream in real time,

· determines the identifier of the sound signal passport.

· receives messages with current coordinates from the Glonass receiver.

From the result of the work, the FFT block forms a packet consisting of a sound passport and transmits it via the CAN bus to the communication controller (CC).

The FFT block is the most loaded sound receiver computer.

The FFT block is implemented on an STM32 microprocessor.

2.3 Communication controller

The communication controller (CC) performs

· receiving recognized sound ID from the FFT block;

· analog-to-digital conversion of temperature sensor readings from other sensors;

· formation of a data package;

· transmission via communications to the signal repository (RS) of a package with signal data.

A packet with signal data can be transmitted to the signal repository in the center of the network-centric bush (Fig. 2) in encrypted form. In the prototype of the CC in the Ethernet local network, the WebSocket mechanism was used as a transport.

In civilian applications, LoraWan radio communications over the MQTT protocol may be a suitable system transport.

Structure of the transmitted signal recording in JSON format:

{ “id”: “sensor identifier”,

“id_sound”: ID,

“db”: “decibels”,

“db”: number,

“x”: number,

“y”: number,

“z”: number,

“timeshtamp”: “time stamp”,

“dt”: “signal duration”,

“dt”: number,

}

If there are appropriate sensors and interfaces in the transmitted record, additional fields are possible, for example,

“tm”: “local temperature”,

“tm”: number,

“w”: “local wind speed”,

“w”: number,

“p”: “local pressure”,

“p”: number,

The CC block is implemented on an STM32 microprocessor.

Currently, a prototype of a multiport CC has been developed using in-line data processing technology and successful tests have been carried out on a personal computer for collecting data and controlling external devices.

3 Description of the composition of the network-centric bush center

The center of the network-centric bush on the right in Fig. 2 consists of two components:

· Signal repository,

· Analyzer.

The above diagram corresponds to the EDA architecture [5].

3.1 Signal repository

The signal repository (RS) of the network-centric system receives from the CC and stores all signals from audio signal receivers in the database.

RS:

· groups signals by time stamps and audio signal passport ID,

· sorts signals by sound strength,

· eliminates friendly sound points,

· collects the required number of observations.

When a sufficient number of observations are collected, the RS transmits the initial data to the analyzer to determine the coordinates of the sound source or generate an alarm.

The source data is transmitted to the analyzer via an Ethernet link.

The PC is implemented on a secure field laptop.

3.2 Analyzer on the central computer side

The analyzer solves a system of quadratic equations and receives the coordinates of the sound source as the roots of the system.

The signal repository and the computer are field computers connected to a network.

The computer can be located on the same laptop as the PC.

4 System functionality

The sound wave receiver consists of a Glonass receiver with an external antenna, an FFT unit with a microphone and a communication controller. The receiver transmits the received and recognized data to the network-centric hive repository.

The interaction of the field sound receiver, signal repository and analyzer is based on the technology of data processing in a stream or event-oriented programming.

The sound wave receiver is a self-contained device and can be installed permanently on the ground, on a floating buoy and/or on a vehicle.

The receivers operate in passive mode and communicate with the signal repository when a corresponding audio signal is recognized. A data transfer session lasts no more than 0.5 mls. A radio beacon mode is possible to quickly find the receiver and return it to the base.

From the “center” commands can be transmitted to test communications, configure the receiver, etc.

The number of receivers in a network-centric system is practically unlimited and is determined by the sufficiency of covering the width of the controlled front or perimeter.

The theoretical required number of receivers depends on the problem conditions:

· to generate an alarm when the equipment operating mode changes or extraneous sound appears – 1 receiver,

· to control the perimeter of a zone with ≥ 2 receivers,

· to determine the sound source on a plane, similar to Benoit’s solution, ≥ 4 receivers,

· to determine 3-dimensional coordinates of a sound source in a quasi-homogeneous medium ≥ 5 receivers.

· to determine 3-dimensional coordinates of a sound source in a heterogeneous environment ≥ 6 receivers.

5 Software implementation platform

It is planned to program software images of receiver microprocessors on a functional programming language platform with subsequent translation in an embedding environment. The software block that receives a signal packet from a WebSocket channel is planned to be implemented in JS. The signal repository and analyzer can be implemented without any exotic language in MS SQL and Python.

6 Estimation of signal propagation time delays

The time to establish the coordinates of the sound source is primarily dictated by the time delay of the arrival of the sound wave front at the receivers. If, for example, the position of the receivers is located so that the front of the sound wave arrives at the last receiver 1 s after the first receiver, then this second will determine the duration of determining the coordinates of the sound source, because For the Python analyzer prototype, the estimated running time of the algorithm was no more than 50ms.

Although the authors of the article are not circuit engineers by training, we considered it necessary to provide rough estimates of the energy consumption, weight and cost of sound receiver components.

7 Estimation of receiver power consumption

The current consumption of the Glonass antenna amplifier device is 32 mA.

Current consumption of the navigation receiver (GeoS-5 RTK) in low power mode ~ 65 mA.

In order to increase the operating time of the receiver from the built-in battery, it is necessary to configure the appropriate operating mode of the loaded FFT microprocessor. The microprocessor is configured to operate with increased performance with a clock frequency of up to 80 MHz with a specific consumption of 120 μA/MHz. Then its current consumption will be 9.6 mA.

The microprocessor of the communication controller is in a standby mode for a long time, followed by a very short period of intensive transmission. Therefore, its energy consumption will be significantly lower, and we will neglect it in calculations of total energy consumption.

The estimated current consumption of the receiver assembly is ~110 mA. In the warm season, when installing an 8000 mAh lithium power supply on board, the continuous operation time of the device will be 72 hours.

8 Receiver weight estimation

9 Estimation of the cost of components of a minimal system

The cost estimate did not include:

· communication equipment,

· central structure of the bush.

conclusions

The proposed set of smart sound receivers as part of a network-centric system is, of course, inferior in its capabilities to the technical means of army artillery reconnaissance presented in [2].

But there are also some advantages. Stationary smart sound receivers are passive, operate in standby mode and practically do not detect themselves.

The system establishes 3-dimensional coordinates of the sound source.

The system is characterized by its low cost, low resource requirements, efficiency and ease of use.

Therefore, the authors of the network-centric system position it almost as a means of individual protection for units up to the level of a battalion or artillery division.

The next article will talk about the tactical locations of smart receivers to cover the front of the battalion and … the positional area of ​​​​the anchorage of a warship.

Literature

1. Artillery sound reconnaissance, ed. MO, M., 1993

2. N.M. PARSHIN, Development and condition of sound means of artillery reconnaissance in the Armed Forces of the Russian Federation, “Military Thought”, 2024, No. 1

3. F. Hueschi, V. Calavri, Stream data processing. – M.: DMK Press, 2021.

4. Ben Stopford, Designing Event-Driven Systems. Concept and design patterns for stream data processing services using Apache Kafka, Irkutsk: ITSumma Press, 2019

5. E.J. Pseltis, Stream data processing. Real-time conveyor.— M.: DMK Press, 2018.