John G. Gruber

Performance Requirements for Integrated Voice/Data Networks

This paper addresses top-down end-to-end user-oriented performance requirements pertaining primarily to voice and digital data services. The discussion of requirements for voice parameters accounts for the performance of existing analog and mixed analog/digital networks, as well as the likely effects on performance of short, medium, and long term evolution toward the ultimate all digital ISDN. The requirements for digital data parameters necessarily reflect an evolutionary process which is less consistent than for voice, and therefore these requirements are less definitive in nature. The discussions of voice and digital data performance apply largely to a wide variety of appropriate network designs, transmission schemes, and switching architectures. Both traditional parameters, as well as contemporary parameters associated with new and evolving systems, are considered. The emphasis is on the performance of nation-wide public and private networks, but the paper also considers the constraints of international connections.

Overview: Performance considerations for integrated voice and data networks

Performance considerations, particularly network delays, for integrated voice and data networks are reviewed. The nature of the delay problem is discussed, followed by a review of concepts, objectives and advances in enhanced circuit, packet and hybrid switching techniques, including fast circuit switching (FCS), virtual circuit switching (VCS), buffered speech interpolation (SI), packetized virtual circuit (PVC), cut-through switching (CTS), composite packets and various frame-management strategies for hybrid switching. In particular, the concept of introducing delay to resolve contention in SI is emphasized and, when applied to both voice talkspurts and data messages, this forms a basis for a relatively new approach to network design called transparent message switching (TMS). This approach and its potential performance advantages are reviewed in terms of various architectural aspects of integrated services networks, such as packet structure, multiplexing scheme, server structure and queuing performance, network topology and network protocols. A number of traffic-management strategies and their grade-of-service implications for voice service are discussed. These strategies include voice call and data session blocking, voice talkspurt and data message buffering, speech loss and data integrity and speech processing techniques, including variable quality, rate, speed and entropy coding. Emphasis is placed on the impact of variable delays on voice traffic, especially the importance of generating and preserving appropriate length speech talkspurts in order to mitigate the effects of variable network delay.

Quality-of-Service in Evolving Telecommunications Networks

This paper addresses the problem of achieving overall quality-of-service of telecommunications networks in North America, following the divestiture of AT&T and the resulting competitive multivendor environment. The paper outlines an approach based on \bullet a top-down process in which quality-of-service is designed into networks by individual network vendors in cooperation with coordinating standards bodies, and \bullet a bottom-up process in which quality-of-service is verified and controlled under actual network operating conditions.

Subjective effects of variable delay and speech loss in dynamically managed voice systems

In this presentation we focus on speech impairments likely to arise in Dynamically Managed Voice Systems (DMVS). Here a DMVS is considered to be a system which may among other things use speech activity detection to exploit speech idle time and manage the resulting talkspurt traffic on a delay and on a loss basis. Examples of such systems include packet voice networks, buffered speech interpolators and advanced integrated services systems. Various speech quality classifications and test criteria used to evaluate the performance of speech systems are reviewed. Emphasis is placed on customer opinion of quality as defined by a rating scale, the objective being to provide the basis for exercising cost‐performance tradeoffs. Based on recent subjective tests grade‐of‐service information is presented pertaining mainly to variable talkspurt delay and to various speech loss mechanisms including front‐end and mid‐talkspurt clipping. The importance of fixed delay in conjunction with echo, echo control and speech detection is also emphasized. Where possible comparisons with previous results are made.

Performance monitoring and surveillance in integrated services networks

The author addresses state-of-the-art monitoring, surveillance, and operational techniques for maintaining digital transmission performance at required levels. Specific areas reviewed are the architectural framework in which operations and maintenance functions are carried out and the transport network elements to be maintained, including physical links and logical paths. Also presented are performance monitoring requirements which should be considered in the development of future industry standards. These requirements include digital transmission performance parameters and associated alert thresholds, as well as local memory requirements, and the modes of reporting performance information across a standard interface to a surveillance center. >

Error Performance Measurements on an Experimental 12/14 GHz Satellite Link

This correspondence presents results related to measurmg the error performance of an experimental 12/14 GHz satellite link established between Carleton University, Ottawa, Canada, and NASA Ames Research Center, California, in 1976-1977. The measurements were performed on a 9.6 kbit/s voiceband data channel which supported the continuous synchronous transmission of fixed length frames (composite packets). The results include measurements of packet error rate versus uplink power settings under various atmospheric conditions, densities of packet error-free interval lengths, within-packet error statistics, and propagation delay variations.

Performance Analysis and Design Guidelines for Multiple Four-Wire Path Telephone Connections

In this paper the listener echo performance of telephone network connections with more than one four-wire path is considered. The objective is to obtain simple design guidelines for the application of four-wire transmission and switching technologies, such that satisfactory listener echo performance for voice and voiceband data services is achieved. The performance analysis is based on a simple but accurate approximation of the overall connection frequency response. The amplitude response of each connection component is uniform. The general case of nonidentical four-wire paths and nonidentical interconnecting twowire paths is considered. Results are presented for: . the overall connection stability margin, and conditions for connection stability, . the condition for satisfactory connection performance, and . the relative merits of allocating required increases in connection stability margin to return losses, and to four-wire and two-wire path losses. A practical example illustrates the results.

Electro‐acoustic factors determining telephone speech quality

In this paper the electro-acoustic factors affecting the quality of telephone speech are categorized as: - listening factors such as loudness, noise, frequency response, and listener echo, - talking factors such as talker echo and sidetone, and - conversational factors such as delay. Listening factors in particular are emphasized and qualitatively interpreted in terms of acoustic pressures at the ear relative to the thresholds of hearing and pain. Typical telephone connections are then considered in which a number of the above factors are manifested; this is done in the context of a network simulation facility used by Bell-Northern Research to simulate various telephone network impairments for the purposes of objective and subjective evaluation. Grade^of-service results that relate customer opinion of transmission quality to the level of a transmission parameter are also discussed. These include listener echo and delay, and subjective-equivalence modelling of speech-correlated digital noise in terms of continuous analog noise.

Estimating the transmission grade-of-service of telephone connections with multiple echo paths

Provides a methodology for estimating the subjective performance of telephone connections with multiple-talker or multiple-listener echo paths. This work is relevant to the performance assessment of voice-grade channels using in-service, nonintrusive measurement devices. The methodology is based on two attributes of existing echo/delay transmission rating models, which enable the dominant echo paths to be identified and all echo paths to be combined into a single path. Thus, as applied to multiple-talker echo paths, the method is more complete than the approach used in the current IEEE voiceband standard. The method, which is illustrated by an example, can also be applied to multiple-listener echo paths. The analysis of multiple echo path problems can be related to balance procedures and to measures of return loss commonly used in the telecommunications industry.<<ETX>>

Ultrasound Cardiogram in Clinical and Physiological Studies

Ultrasound cardiograms have been obtained from over 3000 subjects over the past 6 yr. The accuracy of this method for the assessment of mitral valve disease, as described in earlier reports, has been confirmed in this large patient study. This method of external, safe study has been found to equal cardiac catherization in accuracy when judged by findings at operation. Equally reliable evaluation of tricuspid valve disease has been obtained, and actually found superior to catherization in preoperative assessment. Valve substance, pliability, and mobility can be determined from the ultrasound records. The prediction of valve characteristics determining whether replacement of a valve with a prosthesis is needed, has been quite accurate. This preoperative information is not obtained from catherization studies. The behavior of the mitral valve, accounting for the opening snap in mitral stenosis, and the Austin‐Flint murmur of aortic regurgitation, have been defined by simultaneous record of direct video displa...