GB2463279A

GB2463279A - Wireless computer access system

Info

Publication number: GB2463279A
Application number: GB0816312A
Authority: GB
Inventors: Martin Tomlinson; Marcel Ambroze; Mohammed Zaki Ahmed
Original assignee: Individual
Current assignee: Individual
Priority date: 2008-09-06
Filing date: 2008-09-06
Publication date: 2010-03-10
Also published as: GB0816312D0

Abstract

The wireless communication system which enables a user to speak commands to a remote computer located at an internet access point typically up to 1.km from the user. The system may resemble from the user's viewpoint a conventional, portable radio used with broadcasting, but with an additional microphone input which may be used to select and to control information transmitted from the remote computer as if it were a local radio broadcasting station. Speech recognition is carried out at the remote computer avoiding the complexity and associated power consumption involved. The invention features time expansion of speech and audio information waveforms with transmitter enabling and disabling which has the benefit of increased range for the wireless links between remote computer and user. The invention may be used at any wireless frequency but will find most applications in the unlicenced Instrument Scientific Model (ISM) and Short Range Devices (SRD) frequency bands suitable for short range, low power wireless links, particularly appropriate for use within buildings.

Description

WIRELESS AUDIO INFORMATION SYSTEM USING

SPEECH BASED COMMANDS

Introduction and Background

This invention is concerned with a robust wireless communication system which enables a user to speak commands to a remote computer located at an internet access point typically up to 1km from the user. The system may resemble from the user's viewpoint a conventional, portable radio used with broadcasting, but with an additional microphone input which may be used to select and to control information transmitted from the remote computer as if it were a local radio broadcasting station.

The complex task of speech recognition, necessary in order to interpret speech based commands, is carried out at the remote computer avoiding the complexity and associated power consumption in-volved, simplifying the user equipment. The invention may be used at any wireless frequency but will find most applications in the unlicenced Instrument Scientific Model (ISM) and Short Range Devices (SRD) frequency bands suitable for short range, low power wireless links, particularly appropriate for use within buildings.

Description of the invention

A block schematic of the invention is shown in Figure 1. The user has a two way, full duplex radio which is able to transmit and receive simultaneously to a remote computer. It is envisaged that the wireless frequencies used will be within the unlicenced Instrunient Scientific Model (ISM) and Short Range Devices (SRD) frequency bands but the invention is not restricted to these frequency bands.

The distance between user and remote computer will typically be short range, up to 1km in length but could be longer depending on radio propagation conditions. The user issues commands to the remote computer by speaking into the microphone input as shown in Figure 1. The microphone converts the speech into a speech waveform which is audio filtered, encoded for transmission and transmitted by the 2 way radio (user) via the antenna. After wireless transmission, this is received by the antenna of the 2 way radio (computer). Each encoded speech waveform that is received is input to the remote computer as shown in Figure 1. The computer decodes each encoded speech waveform and carries out speech recognition in order to interpret each received command. The interpreted commands are used by the remote computer to determine the information required by the user. The required information if not already locally stored on the remote computer is fetched via the internet connection shown in Figure 1. The required information if not already in audio format is converted to audio and encoded for transmission by the 2 way radio (computer) via the antenna. The radiated wireless transmission is received by the antenna of the 2 way radio (user).

The received, encoded information is decoded by the 2 way radio (user) and output as sound by means of a loudspeaker, or other audio transducer means as shown in Figure 1.

There are strict limits imposed by the wireless licensing authorities on the maximum radiated power permitted in any given frequency band and this in turn limits the practical range of wireless transmission. It is a fundamental law of information theory [1] that (1) and the information rate, C, able to be communicated by the wireless communication system, is proportional to the bandwidth, W, and the logarithm of the ratio of the received signal power, S plus noise power, N, to the noise power, N. For a given communication technology, the ratio of the received signal power,S plus noise power, N, to the noise power, is a fixed ratio constant Rjec.

Accordingly jçj=Rtecl (2) and C W1og2R (3) with noise spectral density defined as N0, N WN0 and W�= (4) log2 Rtec As ___ (5) it follows that = N0(R1) (6) S > C (7) No(Rtec -1) -1og2Rt and it is found that s �= CN0(R -1) (8) lOg2Rtec The term N0(Rt-1) is a constant for a given communication technology and may be conveniently represented as the constant Mtec. The final result is the relationship between the received power and the information rate able to be communicated, namely S �= CMtec (9) This indicates that the required, received signal power is directly proportional to the information rate carried by the communication system. Since received signal power is proportional to transmitted power it is apparent that the required transmitter power is directly proportional to the information rate carried by the communication system. Correspondingly it is a fundamental law of communica-tions [2] that given the wireless propagation conditions and the required distance between transmitter and receiver, the required transmitter power is proportional to the information bit rate expressed in bits per second. As shown by Shannon [1] all conveyed information, even analogue based infor-mation such as speech, may be expressed in terms of information bit rate. It is a feature of the invention that the practical range of the 2 way wireless transmission link between user and remote computer is extended by time domain scaling of transmitted speech commands and the transmitted audio information, effectively reducing the information bit rate in each direction. Figure 2 shows the arrangement for the transmission and reception of the speech commands from the user for the case where analogue encoding is used. The output of the microphone is firstly amplified and filtered as shown in Figure 2 prior to time expansion. The filtering is chosen such as to compromise between restricting the bandwidth, reducing the information rate after encoding but reducing the fidelity mak-ing speech recognition of each spoken command more prone to error. By way of example bandwidth restriction to 5kHz is considered and the case is considered of the word "command" spoken into the microphone. After amplification and filtering to a bandwidth of 5kHz, the waveform is shown plotted in Figure 3. As can be seen from the waveform the duration of the sound "command" is just over 500mS which is determined by the duration detector shown in Figure 1. The duration of the wave-form needs to be known in order to carry out time expansion. Time expansion of the waveform may bye implemented by the well known technique [2, 3, 4] of time domain sampling, writing to memory, change of sampling clock, reading from memory followed by interpolation filtering. For the example case, the waveform may be sampled at 20kHz, storing approximately 10,000 samples in memory and reading these samples at 5kHz sampling rate causing an expansion of duration to 2 seconds as shown in Figure 4. In the example case of the sound "command", this waveform would be present at the output of the time expansion sub-system shown in Figure 2. The time expanded waveform is encoded by the analogue encoder shown in Figure 2 prior to modulation onto a wireless carrier whose frequency may be selected to be within one of the ISM or SRD unlicensed frequency bands. Alter-natively, if the user has made suitable arrangements with the radio licensing authorities, a licensed frequency band may be used. Any one of a number of well known techniques [2, 4] may be used for analogue encoding and modulation prior to the wireless transmission of the time expanded waveform.

As shown in Figure 5, each wireless transmission is received by the antenna, converted into a waveform and analogue demodulated by the analogue demodulator. The analogue output is decoded by the analogue decoder to produce a received approximation of the time expanded waveform present in the 2 way radio (user). The output waveform is time compressed and input to the remote com-puter to be interpreted as a command from the user by standard speech recognition software. As a further embodiment of the invention, the speech recognition software may be modified to be able to recognise time expanded speech waveforms and the arrangement for this further embodiment is shown in Figure 6.

The degree of time expansion depends upon the acceptable limits of response time imposed by users and the required increase in wireless range. For the example case of the word "command", time expansion by the factor 4, gives an overall response time of typically 2.5 secs and typically an % increase in wireless range.

In a further embodiment of the invention, digital encoding of each speech waveform is carried out as shown in Figure 7. The output of the microphone is amplified and filtered. Time expansion may be carried out followed by digital encoding similar to the analogue encoding case but it is more straightforward to digitally encode each speech waveform followed by time expansion as shown in Figure 7. With time expansion, the duration of each bit is time expanded. After time expansion, the number of bits transmitted for a given speech waveform is the same, but the information rate required to be transmitted by the 2 way radio (user) is reduced. Standard, well known techniques [3] may be used to digitally encode each speech waveform. Lossy compression is commonly used in digital encoding. A well known example is MP3 compression and the degree of compression is traded off against ease of speech recognition by the remote computer. Following time expansion the resulting bit stream is modulated onto a carrier and radiated by the antenna. Standard digital modulation techniques may be used [2, 4] such as Phase Shift Keying (PSK) or Frequency Shift Keying (FSK).

The receiving arrangement is shown in Figure 8. Each wireless transmission is received by the an-tenna, converted into a waveform and demodulated by the digital demodulator. The digital output is time compressed and input to the digital decoder to produce a received approximation of the speech waveform present in the 2 way radio (user) and input to the remote computer to be interpreted as a command from the user by standard speech recognition software.

The audio information transmitted by the remote computer may be processed in a similar manner in order to extend the range of the wireless link from the remote computer to the user. In a further embodiment of the invention as shown in Figure 9, each audio waveform communicating informa-tion to the user, after amplification and filtering is time expanded in the time expansion sub-sytem following duration detection by the duration detector. The time expanded waveform is analogue encoded by the analogue encoder and the output is modulated by the modulator prior to radiation by the antenna as shown in Figure 9.

The corresponding receiving system at the user end of the wireless link is shown in Figure 10.

The signal output from the receiving antenna is analogue demodulated by the analogue demodulator and the output is analogue decoded by the analogue decoder to produce an approximation to the time expanded audio information waveform which was present at the remote computer. The output of the analogue decoder is time compressed by the time compression sub-system and input to the audio interface where the information waveform is converted to audio by an audio transducer such as a loudspeaker or headphones in order to be heard by the user.

In another embodiment of the invention digital encoding and digital modulation may be used instead of analogue processing. As shown in Figure 11, each audio waveform communicating infor-mation to the user, after amplification and filtering is digitally encoded by the digital encoder. The duration of the audio waveform is determined by the duration detector. The output of the duration detector is used by the time expansion sub-sytem to extend the duration of the digitally encoded output and the time expanded bits are modulated by the digital modulator prior to radiation by the antenna as shown in Figure 11.

The corresponding receiving system at the user end of the wireless link is shown in Figure 12. The signal output from the receiving antenna is digitally demodulated by the digital demodulator and is time compressed by the time compression sub-system and input to the digital decoder to produce an approximation to the audio information waveform which was present at the remote computer. The output of the digital decoder is input to the audio interface as shown in Figure 12 where the infor-mation waveform is converted to audio by an audio transducer such as a loudspeaker or headphones in order to be heard by the user.

In a further embodiment of the invention the information about the duration of the transmitted, time expanded waveform is used to switch the transmitter on and off. As an example Figure 13 shows the speech command sub-system with digital encoding and modulation with the added feature of transmitter enabling and disabling. The trasmitter power amplifier is only switched on while the time expanded, digitally encoded speech waveform is present at the input to the modulator. Following the end of the time expanded, digitally encoded speech waveform input to the modulator, the transmitter is disabled, ie switched off. In some frequency bands, particularly the ISM and SRD frequency bands, higher transmitter powers are permitted for low duty cycle transmitters which in turn extends the range of the wireless link between user and remote computer. It is clear that transmitter enabling and disabling may be carried out whether the encoding is analogue or digital and also for wireless transmissions from the remote computer to the user.

References [1] C.E. Shannon, A Mathematical Theory of Communication., Bell System Technical Journal, vol. 27, pp. 379-423, 623-656, July, October, 1948 [2] J.G.Proakis, Digital Communications, McGraw-Hill, 1997 [3] C.Wooton, A Practical Guide to Video and Audio Compression,Elsevier, 2005 [4] L.W. Couch, Digital and Analog Communication Systems, 5th ed., Prentice Hall, 1997