scispace - formally typeset
Search or ask a question

Showing papers on "Packet loss published in 1986"


Journal ArticleDOI
TL;DR: In this article, the authors explore techniques for replacing missing speech with wave-form segments from correctly received packets in order to increase the maximum tolerable missing packet rate in voice communications.
Abstract: Packet communication systems cannot, in general, guarantee accurate and prompt delivery of every packet. The effect of network congestion and transmission impairments on data packets is extended delay; in voice communications these problems lead to lost packets. When some speech packets are not available, the simplest response of a receiving terminal is to substitute silence for the missing speech. Here, we explore techniques for replacing missing speech with wave-form segments from correctly received packets in order to increase the maximum tolerable missing packet rate. After presenting a simple formula for predicting the probability of waveform substitution failure as a function of packet duration and packet loss rate, we introduce two techniques for selecting substitution waveforms. One method is based on pattern matching and the other technique explicitly estimates voicing and pitch. Both approaches achieve substantial improvements in speech quality relative to silence substitution. After waveform substitution, a significant component of the perceived distortion is due to discontinuities at packet boundaries. To reduce this distortion, we introduce a simple smoothing procedure.

247 citations


Journal ArticleDOI
TL;DR: In this article, a simple congestion control scheme using the acknowledgment timeouts as indications of packet loss and congestion is proposed, which can be used in any network with window flow control, e.g., ARPAnet or ISO.
Abstract: During overload, most networks drop packets due to buffer unavailability. The resulting timeouts at the source provide an implicit mechanism to convey congestion signals from the network to the source. On a timeout, a source should not only retransmit the lost packet, but it should also reduce its load on the network. Basedon this realization, we have developed a simple congestion control scheme using the acknowledgment timeouts as indications of packet loss and congestion. This scheme does not require any new message formats, therefore, it can be used in any network with window flow control, e.g., ARPAnet or ISO.

119 citations


Proceedings ArticleDOI
07 Apr 1986
TL;DR: Techniques for replacing missing speech with waveform segments from correctly received packets in order to increase the maximum tolerable missing packet rate are described.
Abstract: Packet communication systems cannot, in general, guarantee accurate and prompt delivery of every packet. The effect of network congestion and transmission impairments on data packets is extended delay; in voice communications these problems lead to lost packets. This paper describes techniques for replacing missing speech with waveform segments from correctly received packets in order to increase the maximum tolerable missing packet rate. After presenting a simple formula for predicting the probability of waveform substitution failure as a function of packet duration and packet loss rate, we introduce two techniques for selecting substitution waveforms. One method is based on pattern matching and the other technique explicitly estimates voicing and pitch. Both approaches achieve substantial improvements in speech quality relative to silence substitution.

69 citations


Patent
08 Aug 1986
TL;DR: In this article, a throughput decision cycle is defined as a certain period reported by a throughput cycle switching report line 51, and the number of packets is counted by a transmission packet (throughput) counter 14 and the counted value is checked by throughput value comparator 16 whether this throughput value exceeds a maximum transmission throughput value stored in a throughput value holding device or not.
Abstract: PURPOSE: To realize the flow control which can cope with the overload condition of traffic, by controlling the transmission throughput by a packet terminal so that its own transmission throughput does not exceed a reported maximum throughput and abandoning excess packets or disconnecting a logical channel by an exchange if the packet terminal transmits packets with a throughput exceeding the reported value. CONSTITUTION: The number of packets is counted by a transmission packet (throughput) counter 14 in every throughput decision cycle which is a certain period reported by a throughput cycle switching report line 51, and the counted value is defined as the transmission throughput in the current cycle and is arranged with a network at the time of originating a call or the like, and it is checked by a throughput value comparator 16 whether this throughput value exceeds a maximum transmission throughput value stored in a throughput value holding device 15 or not. If it does not exceeds, a transmission permission report line 54 is set to the transmittable state and a packet transmission (throughput) controller 12 transmits packets from a transmission packet buffer 13. If it exceeds, the transmission permission report line 54 is set to the untransmittable state and the packet transmission (throughput) controller 12 stops the transmission. COPYRIGHT: (C)1988,JPO&Japio

14 citations


Patent
13 Aug 1986
TL;DR: In this article, the image information is transferred by dividing it into packet units, resending only a wrong packet, and restoring it at every packet, which can effectively utilize network resources such as a line, etc by maintaining the transfer efficiency.
Abstract: PURPOSE: To effectively utilize network resources such as a line, etc by maintaining the transfer efficiency, by transferring the image information by dividing it into packet units, resending only a wrong packet, as for the image information which has caused a packet error or a packet loss, and restoring it at every packet CONSTITUTION: One of image terminal equipments 11 which are connected to a switching network PS is used as an outgoing terminal 11, the other is used as an incoming terminal 11, and when image information is transferred between both terminals, each of information blocks is divided, reconstituted to plural packets 12, and the transfer is executed by a packet unit When a loss or an error of the packet 12 is detected by the incoming terminal 11, the incoming terminal 11 requests to the outgoing terminal 11 to resend the packet 12 which is lost or made an error and executes its restoration, and a transmittal confirmation between both the outgoing and incoming terminals is executed by using an exclusive block for resending the transmittal confirmation, after a transfer of plural packets 12 belonging to one information block is ended In such a way, a high transfer efficiency can be maintained, and network resources such as a line and an exchange, etc can be utilized effectively COPYRIGHT: (C)1988,JPO&Japio

8 citations


Journal ArticleDOI
TL;DR: The proposed protocol provides high-speed transmission function with priority rules, transmission error correcting functions and packet transit delay indicating functions for voice and video communications, in addition to the X. 25 protocol functions for data commmunications.
Abstract: This paper proposes a multi-media packet protocol for realizing packetized data, voice, and video communications in a packet switched network. Requirements for such multi-media packet communications services are greatly manifold, ranging from low to high speed, low to high grade service quality or simple to sophisticated service facilities, depending on the communications media.The proposed protocol provides high-speed transmission function with priority rules, transmission error correcting functions and packet transit delay indicating functions for voice and video communications, in addition to the X. 25 protocol functions for data commmunications. Also, a newlylayered structure for this protocol is presented, in which functional overlaps among layers are eliminated and common functions used by every medium are arranged into lower layers.

4 citations


Journal ArticleDOI
TL;DR: This paper describes a technique for reducing the delay of speech packets on Ethernet caused by collisions with other packets by truncating the standard binary exponential back-off generated by nodes experiencing a collision.
Abstract: This paper describes a technique for reducing the delay of speech packets on Ethernet caused by collisions with other packets. The improvement is achieved by truncating the standard binary exponential back-off generated by nodes experiencing a collision. The technique is easily implemented using special-purpose v.l.s.i. Ethernet devices and ensures an even distribution of delay and packet loss between active nodes.

4 citations


Patent
05 Nov 1986
TL;DR: In this article, the authors proposed to evade the concentration of load on a link in a self-routing network for an information packet and to improve the throughput of the network, by providing an input network controller and an output network controller separately for an input and output part packet handlers.
Abstract: PURPOSE: To evade the concentration of load on a link in a selfrouting network for an information packet and to improve the throughput of the network, by providing an input network controller and an output network controller separately for an input and an output part packet handlers. CONSTITUTION: An inputted information packet is received at the packet handler 11 in the input part of a system, and the signal of link direction information supplied from a second self-routing network 13b is attached on the header of the packet, and it is supplied to a first self-routing network 13a. Also, the output part packet handler 12 outputs the packet supplied through the network l3a to the outside. And the network 13a switches the information packet, and the network 13b recognizes the linking state of the network 13a by the output part network controller 14, and sends a signal to the handler 11 via the input part network controller 15. COPYRIGHT: (C)1988,JPO&Japio

3 citations



Proceedings ArticleDOI
Nachum Shacham1
01 Oct 1986
TL;DR: Adaptive protocols that control transmission parameters such as power, bit rate, code rate, and inter-transmission time can be used to stabilize the network characteristics as seen by high level protocols and to enhance the network performance.
Abstract: The special characteristics of multi-hop packet radio networks result in dynamic channel behavior and time-varying network topology whih cause variations in the signal-to-noise ratio of received packets and in the congestion levels Transmission parameters such as power, bit rate, code rate, and inter-transmission time can be used to stabilize the network characteristics as seen by high level protocols and to enhance the network performance Adaptive protocols that control these parameters are described and their effects on the network performance and their interactions are discussed

1 citations




Proceedings ArticleDOI
01 Dec 1986
TL;DR: This paper presents actual implementations of packet voice communication systems over two types of PC based local area networks, one is a token-passing ring network and the other is an Ethernet network.
Abstract: This paper presents actual implementations of packet voice communication systems over two types of PC based local area networks. One is a token-passing ring network and the other is an Ethernet network. The system configuration, system operation and system performance analysis is described for both networks. A formula for the maximum allowable number of active voice stations is presented for both systems.The last part of the paper describes a proposed design for a distributed packet voice communications protocol. The protocol presented deals with the higher levels of the communication system. The purpose of the protocol is to establish and maintain a telephone conversation between two users by using the underlying network services.