3.3 .Beamforming
Beamforming is spatial filtering exactly like to frequency domain analysis of time signals. It is used to maximize the sound quality. There are many definitions of Beamforming as follow
- Beamforming is a signal processing technique for temporal spatial filtering, that takes inputs from multiple Microphones of the array (often linear) calculate the distance from the sound source by using time delay between sensors.
- Beamforming Is one of the simplest and most robust means of Spatial Filtering, i.e., discriminating between signals based on the physical locations of the signal source.
.There are some technical terms associated with the Beamforming technique
- Directivity
A beam-former is a spatial filter and can be used to increase the signal-to-noise ratio by blocking most of the noise outside the directions of interest. It is a relation between frequency and direction like individual microphone.
- Directivity index (DI)
It is the ratio of power into the given/ looked direction with respect to average power over all
Directions
- Beam steering
A beamformer can be electronically steered/ focused, towards specific direction. The amplitude and phase of each element is controlled
- Beam Pattern
It is the receiving pattern of beamformer where, receiving sensitivity is the function of direction. It is also known as directivity pattern.
3.1. Principle of Beamforming
It works on the linearly combination of the individual microphone output signals for:
- Enhancement the signal arriving from a given/looked direction
- Suppress the signal coming from unwanted directions
Mathematically, it can be written as
Each sensor output is multiplied with weight factor before summing up/
There are various types of Beamformer and each has its own characteristics:
- Delay-and-Sum Beamformer
- Filter-and-Sum Beamformer
- Minimum Variance Distortion-less Response Beamformer (MVDR)
4. Design of Acoustic Measurement System
The design of the Microphone arrays for Acoustic Measurement System is very critical process because to achieve desired performance involves trade-offs among the Array geometry, sensors count, signal-to-noise ratio (SNR), as well as many other factors. In general, the following aspects are considered
- Characteristics of the Microphone Array geometry for achieving good sound capturing
- Selection and placement of the appropriate number of Microphones related to application
- Design requirements for low noise, directionality and microphones sensitivity matching
Requirement Specification
The basic specification and geometry of the Microphone Arrays for speech localization and steered response through Beamforming techniques like Delay and Sum, Filter and Sum or MVDR at a frequency of 100-1500 Hz is given as
Array Type | Number of Elements | Spacing | Geometry |
Linear | 8 | Uniform | |
Linear | 8 | Non Uniform | |
Circular | 8 | Uniform | |
4.1.Design Specification
A customized and state-of-art hardware design is proposed after considering different possible scenarios, processing capability, future needs and also like;
- Standalone Processing or Host-PC based Processing
- Upgradeable with future coming advance Embedded processors
- Expandable with other sensors to perform Sensor Fusion
The core building blocks of hardware are
- Digital MEMS Microphone
- Microprocessor
- Communication Interfaces and Mass Storage
- Expansion Connectors