WO2000052441A1

WO2000052441A1 - Method and apparatus for determining the perceptual quality of speech in a communications network

Info

Publication number: WO2000052441A1
Application number: PCT/US2000/005104
Authority: WO
Inventors: Lewis L. ROBERTS, Jr.; Bassem Rustom; John David SHAW
Original assignee: American Towers, Inc.
Priority date: 1999-03-04
Filing date: 2000-02-28
Publication date: 2000-09-08
Also published as: AU4169200A

Abstract

A system and method for providing an objective speech quality evaluation of a mobile telecommunications network. It includes a first transceiver (1) with memory configured to store a speech recording, and configured to transmit the speech recording over a communication network. It further includes a second transceiver (113) configured to receive the sample recording from the first transceiver (101) and an objective speech quality analyzer coupled to the second transceiver (113) and configured to measure the objective speech quality of the received speech recording in real time. At least one of the first transceiver (101) and the second transceiver (113) is configured to be mobile.

Description

METHOD AND APPARATUS FOR DETERMINING THE PERCEPTUAL

QUALITY OF SPEECH IN A COMMUNICATIONS NETWORK

FIELD OF THE INVENTION

This invention relates to the art of evaluating the quality of speech transmitted over

telecommunication networks and, more specifically, evaluating the subjective quality of transmitted speech using an objective method.

BACKGROUND OF THE INVENTION

In designing, deploying and maintaining wireless communication networks different criteria are taken into consideration with the ultimate goal of providing high quality speech

transmission over the networks so as to achieve high listener satisfaction. Listener

satisfaction is inversely proportional to the amount of distortion introduced by the

transmission medium in a communications network. In order to predict and achieve high

listener satisfaction, a reliable and cost-effective method of measuring the quality of the

transmitted speech is required. By assessing the results of the measurement, a service

provider or an operating telecommunications company can make adjustments to the network

configuration, if needed. A number of measurement systems to measure the quality of speech

have been proposed for this purpose. These systems can be divided into two broad

categories: subjective and objective speech quality measurement systems.

Subjective quality measurement systems conduct surveys with human respondents and

generate reports based on the responses given by the subjective assessment of the human

respondents. Because the goal of speech quality measurement techniques is to predict how well a human listener is satisfied by the quality of the speech transmitted by the wireless

medium, human surveys appear to be a reasonable choice. However, human trials and

surveys are expensive to conduct because of the time and effort of surveyors, which include

careful planning prior to conducting the surveys, supervising the surveys, and analyzing

collected survey data. In addition to the cost of the surveyors' time, human respondents

require recruitment and pay for the time they spend in the surveys. Such costs can mount

very quickly and are often perceived as exceeding the value of speech quality assessment. In

a typical scenario, tests conducted for one cellular location may require recruitment of 20-100

human testers, which could cost $20,000 per survey. Furthermore, such tests require the

transmitted speech to be recorded at receiver sites, the recordings transported from the

receiver sites and analyzed by respondents at later time and at a central office. These steps

induce additional delay to the surveys.

Due to these and other shortcomings, objective methods of measuring human

satisfaction have been devised to replace the subjective methods. An objective speech

quality measurement system is described in U.S. Patent No. 4,860,360, which describes a

system of evaluating the quality of speech in a sample communication system in a speech

processor. This patent discloses a method wherein an undistorted sample of speech is passed

through a set of critical band filters to provide a first power spectrum. The undistorted sample of speech signal is then transmitted via a transmission medium and received as

received speech. The received speech, which could be distorted by the transmission medium,

is also passed through the set of critical band filters to provide a second power spectrum. A

variance-covariance matrix is calculated from the first and the second power spectra. Finally,

a Mahalanobis D2 calculation is performed on the matrix, yielding D2 data, which represents a numerical value representing the quality of speech in the sample file. Several other methods to measure the transmission quality of telephone speech are suggested in the P-Series ITU-T recommendation, P.861, entitled "Objective Quality Measurement of Telephone Band (300-3400 Hz) Speech Codecs." The recommendation suggests a method of producing source speech for objective measurement, calculating an objective quality based on an algorithm for measuring objective quality of speech called the

Perceptual Speech Quality Measure (PSQM), and estimating from the objectively measured quality coefficients the subjective equivalent values for comparison.

U.S. Pat. No. 5,095,500 by Tayloe et al., (the Tayloe patent) discloses a method and system of evaluating radio coverage of a geographic area serviced by a digital cellular radio telephone communication system comprising a plurality of base stations each having a transmitter and a receiver and a plurality of mobile units having co-located transmitters and receivers for transmitting and receiving communication message signals between the base stations and a mobile unit. The object of the Tayloe patent is to obtain signal strength and signal quality of transmissions from a serving and adjacent base stations, to report this data to the serving base stations, and to permit the serving base station to regulate the process of switching the established call from one cell to another as the mobile unit travels from cell to cell. The position of the mobile unit is determined when a call is received at a base station in order to pinpoint the area within a cell that possesses the electromagnetic characteristics revealed by the data gathered by the mobile unit. The Tayloe patent describes a method of obtaining certain transmission parameters from a call made by a mobile unit and does not disclose or claim to measure the perceptual quality of speech in a digital cellular telephone system. It should be noted that a "call" initiated according to this patent does not include a transmission or reception of spoken utterances. Moreover the data gathered according to this invention is evaluated at the base station. U.S. Pat. No. 5,481,588 by Rickli et al. (the Rickli patent) discloses and claims a method of obtaining data on the quality of service of a radio telephone system within a coverage area. A mobile unit is equipped to make telephone calls, acquire and store call identification and connection parameters and the time and position data. Similar to the Tayloe patent, the Rickli patent is not directed to measuring the perceptual quality of speech signals transmitted over the radio network. The understandability of the speech signals transmitted over the radio network was not disclosed or claimed by this patent.

Thus, none of the methods of evaluating a communications network that are currently in use have used an objective method of analyzing the quality of speech signals in a communication network to evaluate the perceptual quality — which is essentially a subjective measure — of human speech in a wireless communications network in real-time conditions. There is a need, therefore, for a system that incorporates an objective measure of calculating the quality of speech in a transmission medium in real time and enable a field operator to make decisions based on a limited set of experiments.

SUMMARY OF THE INVENTION The present invention is directed towards a method and apparatus for measuring the subjective quality of speech using an objective method in a communication network. In one aspect, the invention comprises a first transceiver is located at a predetermined location; and a second transceiver is located at a different location, preferably mounted to a mobile unit. In case the second transceiver is mounted to a mobile unit, the mobile unit roams along a route — which is either pre-determined or arbitrary — in a geographical area. Samples of speech signals obtained from spoken phrases are stored in both the first and the second transceivers. At a scheduled time, either the first transceiver or the second transceiver initiates a telecommunication connection with the other transceiver and engages in a communication by transmitting or receiving or simultaneously transmitting and receiving designated speech signals from the stored speech samples. At each of the first and the second transceivers, the received speech signals are compared with stored speech signals to develop a numerical measure representative of the quality of received speech. In another aspect, the present invention also includes a Global Positioning System

(GPS) coupled to the second transceiver which determines the position of the mobile unit at predetermined time intervals and records the route as well as the numerical measure representing the quality of the speech in a report format. In another aspect, the GPS is used for precisely synchronizing the transceivers. In a yet another aspect, the invention is directed to obtaining transmission parameters using an air interface device coupled to the first or the second transceiver, or both transceivers. In a further aspect, the invented method comprises correlating the transmission parameters with the numerical measure representative of the quality of received speech. In another aspect, the invention is computer executable software code stored on a computer readable medium configured to perform the steps involved in the method above. In a further aspect, the invention is a programmed computer executing the software code.

BRIEF DESCRIPTION OF THE DRAWINGS These and other objects, features and advantages of the invention will be more readily apparent from the following detailed description of a preferred embodiment in which:

FIG. 1 is a schematic depicting the arrangement of the apparatus according to a preferred embodiment of the present invention;

FIG. 2 is a flow chart describing the steps in a preferred embodiment of the present invention; and FIG. 3 depicts an illustrative evaluation area with a plurality of evaluation routes. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to FIG. 1, a first transceiver 101 is connected to a first communication network 107 via a first interface device 103. A second transceiver 113 is connected to a second communication network 108 via a second interface device 109 and a mobile handset 124. The first and the second communication networks, 107 and 108 are interconnected.

Preferably the second communication network 108 is a wireless network. In an alternative embodiment, the first communication network 107 is a wireless network. In other embodiments the communication networks 107 and 108 could be any other networks, such as the public data networks or the Internet or any other packet or circuit switched network or combinations of networks. Though networks 107 and 108 are shown, the principles of the present invention do not require a networked interconnection between the first and the second transceivers. They could be connected via a broadcast, point-to-point wireless, or wireline connection or any other method.

Advantageously, the second interface device 109 is an air interface device such as a wideband air interface device based on W-CDMA and TD-CDMA technologies. A formal definition of the air interface standard for cellular telephones can be obtained from the standard TIA/EIA/IS-95-A, entitled "Mobile Station - Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System." A definition of air interface for Personal Communications Service (PCS) at frequencies in the range of 1850-1990 Mhz can be obtained from the standard ANSI J-STD-008, entitled "Personal Station - Base Station

Compatibility Standard for Dual-Mode Wideband Spread Spectrum PCS System." Examples of commercially available air interfaces are those marketed by wireless communication equipment manufacturers such as Nortel®, Alcatel®, Nokia®, Motorola®, Lucent® and others. This configuration allows the second transceiver 113 to be mounted to the vehicle and roam the evaluation area. The air interface device 109 provides a way for the second transceiver 113 to communicate with the mobile handset 124 using standard signaling methods. The signals transmitted and received by the air interface device 109 can be intercepted by the second transceiver 113. These signals indicate the state of the connection between the first and the second transceivers. Thus, by intercepting and decoding the air interface messages, accurate determination of transmission parameters can be made. In a preferred embodiment, these transmission parameters are chosen from the group comprising the following: timing alignment, transmit power, radio frequency (RF) channel, received signal strength, bit error rate, frame error rate, aggregate pilot, dominant pilot, signal quality error (SQE), R-XQUAL (used in relation to the Global System for Mobile— GSM— Standard) and other air link data analysis parameters such as forward and reverse link data analysis.

Using these parameter values, it is possible to establish a correlation between the signal transmission parameters and the perceptual quality of speech and to obtain a customized evaluation of the quality of reception of wireless speech.

It should be noted that the present invention is directed toward measurement of the quality of speech in a wireless network comprising a cellular network or a Personal

Communications Service (PCS) network or any other similar network.

The first transceiver 101 preferably is equipped with memory 115, such as a semiconductor memory, a disk drive or a portable medium such as a -floppy disk, and a first processor 102, such as a Pentium IF^M microprocessor. The first processor 102 is preferably operable at clock speeds greater than or equal to 300 MHz. Processor clock speed at or above

300 MHz enables the analysis of speech in real time. The first transceiver 101 is equipped with a telephony interface card to communicate with the telephone network, commonly called the Public Switched Telephone Network (PSTN). Though a PSTN is described in a preferred embodiment, the invention contemplates other networks as well. For example any communications network such as a data communication network or a packet switched

- 1 - network is also envisioned to work with the same inventive principles. The first transceiver

101 also includes an electronic circuit 117 to perform objective analysis of speech and data files 118, which are stored in the memory 115 coupled to the first transceiver 101. The electronic circuit 117 is configured to perform as an Objective Speech Quality -Analyzer (OSQA).

The second transceiver 113 comprises a processor, such as a Pentium IF^M microprocessor, coupled to which is a memory device 119, such as a semiconductor memory, a disk drive or a portable medium such as a floppy disk. The memory device 119 is coupled to an OSQA 121. Optionally, the second transceiver 113 is coupled to a digital-to-analog conversion device 122 to convert speech from digital to analog signals and vice versa. The digital-to-analog conversion device 122 is then connected to a mobile telephone handset 124 via a pair of leads 126. It should be noted that the digital-to-analog conversion device is not required if the speech signals are already digitized when presented to a transceiver. The mobile telephone handset 124 is connected to the telecommunication network 107 via a wireless telecommunication network 108. The second transceiver 113 is preferably coupled to a mobile unit 130, which allows mobility to the second transceiver 113. Examples of the mobile unit 130 comprise an automobile or other vehicle, a bicycle or other manually movable object, or a backpack attached to a walking human. In a preferred embodiment, the second transceiver 113 is mounted to a vehicle such as a van. Both the first transceiver 101 and the second transceiver 113 store in the associated memory devices 115 and 119 respectively, copies of pre-recorded speech files containing speech utterances by several test speakers. These speech files are preferably recorded in a studio environment so as to limit the amount of ambient noise added to the signals and are used as reference signals containing undistorted speech recordings. Each of the several pre- recorded speech files is associated with an identifier. The first transceiver 101 and the second transceiver 113 also are equipped with scheduling devices containing schedule for transmission and reception of the reference speech files at precise times. In a preferred embodiment, the scheduling device is a computer file containing the time schedule in a format accessible by the first transceiver 101 and the second transceiver 113. The second transceiver 113 is coupled to a Global Positioning System (GPS) device

127. The GPS device 127 provides the location information of the second transceiver 113 — the latitude and the longitude — periodically as the mobile unit 130 to which the second transceiver 113 is coupled, moves. The second transceiver 113 is also coupled to an information system 128 that is designed to work with data referenced by spatial or geographic coordinates, called the Geographical Information System ("GIS").

The GIS is both a database system with specific capabilities for spatially-referenced data, as well as a set of operations for working with the data. The GIS enables a user to identify in a two- or a three-dimensional map a location and thereby helps specify, describe and configure the geographical area in which the quality of wireless speech is measured by the present invention. The GIS package used is preferably Geographic Resources Analysis

Support System (GRASS)™, available from Baylor University, Waco, Tx., or Arc/Info™, available from Redlands, Ca.

The GIS and the GPS systems function in the following way. The GPS generally is composed of a transponder mounted on a vehicle that communicates with at least one geostationary satellite orbiting the Earth. At pre-determined periods, the transponder sends packets of data to the satellite with which it communicates. The satellite will determine precisely the latitude and the longitude of the vehicle and returns this data in a data packet in a response message to the vehicle. Thus, the GPS enables the precise determination of the location of roaming vehicle at pre-determined intervals of time. The GIS is a mapping program that enables a user to plot the location of the vehicle as it roams along a route. Thus, the precise location is obtained by the GPS, and the plotting of the route is accomplished using the GIS software. When the two systems work together, a recording of the vehicle location in a graphical manner is accomplished. It should be noted that the invention does not contemplate the GIS and the GPS systems as essential components of the apparatus. Therefore, the invention should not be limited to being inclusive of these two components.

Conceptually, there are two components that comprise an objective speech quality analyzer: the front-end digital signal processor (FE-DSP), and the speech quality analyzer. These two components can be implemented in many ways. For example, the OSQA can be an electronic circuit 117 configured to evaluate the objective quality of speech. Alternatively, the OSQA can be implemented entirely in software as a computer-executable program code running on a general purpose computer such as a personal computer comprising a processor such as a Pentium II™ microprocessor, a storage device such as a disk drive and a memory such as a semiconductor memory. It should also be noted that a part or the whole of the circuit 117 comprising the OSQA can be implemented in hardware in order to speed up the time taken for computation. For example, the FE-DSP circuit could be implemented by programming a general purpose signal processor such as the TMS-C30™ Processor marketed by Texas Instruments, Inc. In an alternative embodiment, the FE-DSP and the OSQA program could be designed to work with an Antares™ circuit board marketed by the Dialogic Corporation, of Parsippany, New Jersey. It has been found that Dual Tone Multi-Frequency (DTMF) signaling can be used to schedule the transmission of the pre-recorded speech files. In that case, a DTMF tone is sent before a particular speech file is transmitted by a transmitting transceiver. However, since the frequencies that comprise DTMF signaling are in the same range as human speech which is the subject matter of this invention, it is possible that DTMF is not unmistakably received and interpreted at locations where reception of signals in the range 300-3300 Hz is unacceptably low. A precise schedule that indicates when a particular speech file is to be transmitted or received overcomes this problem for areas where the quality of received signal in that frequency range is below an acceptable threshold. Thus, in a preferred embodiment of the present invention, a time-schedule is utilized. It must be noted, however, that the invention contemplates a signaling method such as an out-of-band signaling method as a way to indicate to the receiver which pre-recorded speech file is being transmitted at any given time.

As an example, the following could be a transmission and reception schedule for a single transceiver 113. Assume that A, B, C, D, E are five pre-recorded speech files stored on a storage device coupled to the second transceiver 113. This ensures that the testing conditions are controlled tightly, thereby resulting in results that are comparable to other similarly obtained results.

It should be noted that there could be other schedules than the one described above. For example, there a the schedules could be designed such that both transceivers must be simultaneously transmitting and receiving speech signals but the speech signals transmitted and received could be different. Additionally, the second transceiver 113 may initiate a telephone call while the mobile unit is in motion or located at a stationary place. In a preferred embodiment, both the first 101 and the second transceiver 1 13 transmit and simultaneously receive the pre-scheduled speech files. In an alternative embodiment, the transmission and reception of speech files, or parts thereof, may occur while the mobile unit is in motion or stationary. The evaluation route 305 along which the mobile unit travels may be a pre-determined route or it may involve an arbitrary path. Referring now to Fig. 3, the second transceiver 113 roams in the network evaluation area 301, which includes a plurality of evaluation points such as station 303 or along an evaluation route 305. Preferably, the second transceiver 113 initiates a mobile telephone call by dialing a telephone number pertaining to the first transceiver 101 located at a station 303. In an alternative embodiment, the first transceiver 101 initiates the call. When the first transceiver 101 "answers" the call by sending an appropriate signal, a two-way communication channel is established.

The first transceiver 101 and the second transceiver 113 are configured to operate together. Once the two-way connection is established between the two transceivers, depending on the pre-programmed schedule, a pre-determined speech signal is transmitted by each of the two transceivers 101 and 113. At the same time, each of the two transceivers is configured to receive the pre-determined speech files over the two-way telephone connection. Once received, the pre-determined speech files are stored as received speech files.

Because both transceivers 101 and 113 operate according to a pre-determined schedule of which speech signal to transmit, it can be determined at both transceivers 101 and 113 which signal is received at any given time. This information is used to retrieve a corresponding stored speech signal and make a comparison between the received and the stored speech files. This comparison is made in real-time either as soon as the received speech signal is received, or while it is received. In a preferred embodiment, this comparison is made by inputting the received speech signal and the stored speech signal to a OSQA device or a computer configured to perform the evaluation. The OSQA device compares the two signals and generates a numerical result called a "score." The score is a measure of the perceptual quality of the received speech signal when the vehicle roams along the evaluation route 305 or located at the station 303. The score is then recorded in the memory devices 115 and 119 associated with the first transceiver 101 and the second transceiver 113 respectively. Also recorded in the second transceiver 113 is the position of station 303 of the vehicle obtained from the GPS device 127 coupled to the second transceiver 113. In an alternative embodiment, the location of the vehicle at the time of testing as the vehicle roams along the evaluation route 305 is also entered by obtaining information from the GPS at several points along the evaluation route 305. It should be noted that the mobile unit may be stationary, or in motion or partly stationary and partly in motion when the transmission or reception of speech is made.

Referring now to Fig. 2, in the preferred embodiment, a set of pre-recorded speech files is obtained. (Step 201). The pre-recorded speech files are stored in the memory 115 and 119 coupled to the first transceiver 101 and the second transceiver 113 respectively (step 203). Preferably the pre-recorded speech files are speech samples marketed by Dynastat

Corporation, of Austin, Texas.

The second transceiver 113 initiates a telephone connection with the first transceiver 101 from a mobile location. Depending on the time at which the telephone connection is made, the scheduled speech recordings stored at the first transceiver are transmitted to the second transceiver 113. The second transmitter 113 also transmits the speech recordings stored at the second transceiver 113 according to a pre-determined schedule. (Step 205). At each of the two transceivers 101 and 113, the speech recordings sent by the other transceiver are received. (Step 207). Subsequently, the received signals are time-aligned with the stored speech files corresponding to the received speech files at each of the two transceivers. (Step 209). During the time-alignment step, if the received speech signal is an analog signal, it is passed through an analog-to-digital (A/D) converter creating a digital file. As noted earlier, if the received speech signal is already digitized, there is no need for the A D conversion step. The digital file is then divided into frames of speech, with each frame corresponding to preferably 10 milliseconds of speech. Of these frames, every tenth frame is selected and analyzed using the OQS A device to obtain a score, indicating the quality of speech signal present in the frame. If the score is greater than a pre-determined threshold value, then the frame is deemed to contain audible speech signal. This is then determined to be the starting point of the speech contained in the received signal. The stored speech signal is then aligned to correspond with the starting point of the speech contained in the received signal. The selection of every tenth frame to perform the detection of high-quality speech signal is a preferred method. In other embodiments, frames can be selected based on other criteria as well.

The OSQA is preferably an implementation of the objective analysis algorithm described in ITU-T P.861. One of the several methods devised to implement the ITU-T

P.861 recommendation is disclosed in WO 96/28950. That application, therefore, is incorporated herein to the extent necessary to understand the principles of the present invention. It is believed that the PSQM algorithm provides an acceptable measure of the perceptive quality of speech files transmitted over a communications network. Additional optimization of the algorithm is performed by incorporating an object-oriented programming structure for efficiency, maintainability and other benefits, and modifying a front-end processor to accept input speech files in a ".wav" format instead of Pulse Code Modulation (PCM) files. It should be noted, however, the method invented contemplates other ways to analyze and determine the perceptive quality of speech, such as Code Excited Linear Prediction (CELP), Linear Predictive Coefficients (LPC), LPC Cepstrum Distance Measure,

Signal-to-Noise Ratio (SNR), or Mean Opinion Score (MOS).

After time-aligning the speech files, the entire received speech signal is applied to the OSQA (step 210), thereby generating a score, which is a numerical value representing the quality of the speech signal received. The present invention is designed to evaluate the quality of received speech along the evaluation route 305 of a pre-designated geographical area called a network evaluation area 301. Alternatively, the network evaluation area 301 could be arbitrarily defined, rather than a pre-designated area. Scores are obtained continuously as the mobile unit roams the evaluation route 305, with the mobile unit either in motion at the time of evaluation or situated at a stationary location along the evaluation route

305.

In a preferred embodiment, the scores obtained by the OSQA are stored in the GIS 128 in order to associate a score with the geographical location where the score is obtained (step 211). After the scores are collected at different points along the evaluation route 305 or a plurality of locations such as station 303, they are used to generate a report summarizing the average score obtained for a route, and the distribution of scores. Preferably, a graphical or tabular report containing the route — along which objective speech quality is measured — together with the scores obtained for each point of measurement is created. Referring back to Fig. 1, in a preferred embodiment, the first interface device 103 coupled to the first transceiver 101 is connected to a wireline telecommunication network, such as the Public Switched Telephone Network (PSTN), while the digital-to-analog conversion device 122 of the second transceiver 113 is connected to a wireless telecommunication network 108 via the second interface device 109. It should be noted that the first interface device 103 coupled to the first transceiver

101 can also be configured to connect to a wireless telecommunication network in a similar manner.

In a preferred embodiment, the networks 107 and 108 — to which the first and the second transceivers are connected — are operated by the same company. In this configuration, the quality of the combination of networks (comprising 107 and 108) is evaluated by measuring the objective quality of the speech received at either the first transceiver 101 or the second transceiver 113. In an alternative embodiment, the two networks 107 and 108 could be owned or operated by different companies. It should also be understood that various modifications will be readily apparent to those skilled in the art without departing from the scope and spirit of the invention. For example, though the above embodiments are illustrated with only two transceivers, utilizing a plurality of first transceivers and a plurality of second transceivers is within the scope of this invention.

Claims

THE CLAIMSWhat is claimed is:

1. An apparatus to measure quality of speech transmitted over a communication network comprising: a first transceiver configured to interface with the communication network; a second transceiver configured to interface with the communication network; and a first objective speech quality analyzer (OSQA) coupled to the second transceiver, wherein the first transceiver is configured to transmit a speech signal which is received by the second transceiver, and the second transceiver is configured to provide said speech signal to the first OSQA, and the OSQA is configured to measure a perceptive quality of the received speech signal in real time.

2. The apparatus according to claim 1, at least one of the first transceiver and the second transceiver is connected to a public switched telecommunication network.

3. The apparatus according to claim 1, at least one of the first transceiver and the second transceiver is connected to a wireless communication network.

4. The apparatus of claim 1, further comprising: a second OSQA coupled to the first transceiver configured to measure the quality of speech signals received at the first transceiver.

5. The apparatus according to claim 4, wherein at least one of the first transceiver and the second transceiver is located in a mobile unit.

6. The apparatus according to claim 5, wherein the mobile unit roams in a predetermined area, which includes a plurality of predesignated locations at which at least one of the first objective speech quality analyzer and the second objective speech quality analyzer is configured to evaluate the speech signals received over the communications network.

7. The apparatus according to claim 6, further including: a geographical information system coupled to the analyzer and configured to record results of the objective quality speech measurements by the analyzer along with the predesignated location information.

8. The apparatus according to claim 7, wherein at least one of the first and the second OSQA includes a perceptual speech quality measurement algorithm.

9. The apparatus according to claim 1, further comprising: an air interface device coupled to the second transceiver and configured to obtain at least one of a plurality of transmission parameters.

10. The apparatus of claim 9, wherein the at least one of a plurality of transmission parameters is selected from the group consisting of timing alignment, transmit power, radio frequency (RF) channel, received signal strength, bit error rate, frame error rate, aggregate pilot, dominant pilot, signal quality error, and R-XQUAL.

11. The apparatus of claim 9 wherein the at least one of a plurality of transmission parameters is stored in a memory device coupled to the second transceiver.

12. The apparatus of claim 9 wherein the transmission parameter is used to correlate with a score representing a perceptive quality of speech.

13. A method of measuring quality of speech transmitted over a communication network, the method comprising the steps of: transmitting a first speech signal over the communication network; receiving said first speech signal to produce a received speech signal; providing said received speech signal to an objective speech quality analyzer (OSQA); comparing in real time the received speech signal with the first speech signal using said OSQA; and obtaining a score representing a quality of speech.

14. The method according to claim 13, wherein at least one of the steps of transmitting, receiving, providing, comparing and obtaining is repeatedly performed at a plurality of locations within a geographical area.

15. The method of claim 13 wherein the score represents a subjective quality of speech.

16. The method according to claim 13, wherein the OSQA uses the ITU-T P.861 algorithm.

17. The method according to claim 14, wherein said plurality of locations are situated along a predetermined route.

18. The method according to claim 17, wherein the first speech signal is transmitted from or received by a mobile unit when it is in motion.

19. The method according to claim 18, wherein the first speech signal is transmitted from or received by a mobile unit is stationary.

20. Computer executable program code stored on a computer readable medium, the code to implement the method of claim 13.

21. A programmed computer configured to execute the code of claim 20.