RU2136113C1 - Method for packet exchange in local-area network - Google Patents

Abstract

FIELD: information exchange between logical objects in local-area networks. SUBSTANCE: method involves relative synchronization of workstations, splitting real time into segments of equal duration each of which is equal to T_c>= t_sp+t_rg,, where b_sp is signal spreading period between two most remote workstations in LAN, t_rg is admissible period for recognition of carrier from other stations in shared channel. In case if this is not detected at station for its interval T_c for information transmission from other stations in network and information for transmission exists, it is transmitted using constant length packets so that beginning of each packet is equal to beginning of packet which is following after "own" segment. Transmission of packet should be finished during interval which duration is equal to T_p= T_cN, where N is number of workstations in this network. Said method for packet exchange can be implemented also for T_p= T_cNk, where k=1, 2, 3,... EFFECT: increased throughput of shared channel for low traffic (under condition that number of operating workstations is in range of 1 to N-1). 2 cl, 8 dwg

Description

 The invention relates to the transmission of information between logical entities, for example, in local networks, in particular, to methods of packet exchange in shared channels.

 In local networks, a significant number of subscriber stations can be connected to a shared channel. A prerequisite for the effective use of the bandwidth of the collective channel in the interests of all subscribers of the network is to ensure the alternate transmission of information from those subscriber stations that have information for transmission. It must be borne in mind that if some form of output of subscriber stations to the collective channel is necessary for packet transmission, then all stations constantly receive information in the network.

 The method of packet exchange determines the functioning of subscriber stations to ensure the transmission of finite-length packets on a shared channel between subscriber stations connected to this shared channel, optimizes the efficiency of using the bandwidth of the shared channel in the interests of all subscribers of this channel.

 A known method of exchange with a temporary separation of the channel of collective use (see "Data transmission systems and computer networks" M., "Mir", 1974, p. 60 [1].

 In the known method [1], the transmission of information in a shared channel from each subscriber station is allowed in a strictly fixed time interval for this station.

 The disadvantage of this method should be attributed to the inefficient use of the bandwidth of the channel of collective use with low traffic subscribers.

 A known method of transmitting messages with random access of subscribers to the channel of collective use (see Shcherbo V.K., Kirichev V.M., Samoilenko S.I. "Standards on local area networks", M., "Radio and communication", 1990 , p. [2].

 The known method [2] for transmitting messages is that all stations collectively use a common physical environment, at each station that has data for transmission, they monitor the state of the physical environment and, in the absence of transmissions from other stations, place their message in the physical environment. If, after the start of transmission, the message in the channel collides with the message of the other station, then each of these stations provides a random combination to the physical environment, recognized at each station as false information. The station then waits for a certain, dynamically changing period of time before retrying the transmission.

 The disadvantages of this method [2] include the fact that when it is used, there is a possibility of simultaneous transmission of several subscriber stations (collisions), which leads to loss of information, delays in its transmission. This probability is higher, the higher the intensity of information transfer in local networks.

 The closest to the proposed facility for its intended purpose, technical nature and the achieved result is a method of transmitting packets in a segmented ALOHA system (see TIIER, M., Mir, Volume 75, N 1, 1987, p. 167 [3].

 In accordance with this method, all subscriber stations are synchronized with each other, the current time is divided into segments of equal duration, if there is information, packets are transmitted in segments allocated for each particular packet station, synchronously so that the beginning of the packet coincides with the beginning of the segment, and the packet is transmitted completely ends during the segment, receive information at all packet stations.

 The disadvantage of this method should be attributed to the inefficient use of the bandwidth of the channel of collective use with low traffic subscribers. The fact is that time segments belonging to subscriber stations that do not have information for transmission remain unused during the entire period of lack of information at these stations.

 The aim of the invention is to increase the efficiency of using the bandwidth of the channel of collective use at low traffic. To achieve the objective of the invention, the problem of ensuring the transmission of information in the segments of those subscriber stations that do not transmit information at present and the uniform distribution of segments of the shared channel between subscriber stations having information for transmission is solved.

The invention consists in the fact that to solve the problem at the subscriber station network
- subscriber stations synchronize;
- the current time is divided into segments of equal duration T _c (T _c ≥ t _p. + t _op ., where t _p. is the propagation time of signals between the two most remote subscriber stations in the local network, t _op. is the time of recognizing the presence of a carrier from others stations in the shared channel, adopted in this local network);
- Assign segments to subscriber stations;
- at each subscriber station, the channel occupancy on “its” time segment is determined by analyzing the carrier during this interval;
- provided that at the subscriber station during its interval T _{c there is} no transmission of information from other subscriber stations of the network and the station has information for transmission, it is transmitted in packets of the same length (T _p ) synchronously so that the beginning of the packet coincides with the beginning of the next for "their" segment, and the transmission of the packet completely ended within the allotted time interval, the value of which is defined as T _p = T _c N, where
N is the number of packet stations for this network.

K - the signs that distinguish the proposed method from the prototype include the fact that
- at each packet station, the channel occupancy is determined in "its" time segment by analyzing the carrier during this interval;
- provided that at the subscriber station during its interval T _{c there is} no transmission of information from other subscriber stations of the network and the station has information for transmission, it is transmitted in packets of the same length (T _p ) synchronously so that the beginning of the packet coincides with the beginning of the next for "their" segment, and the transmission of the packet completely ended within the allotted time interval, the value of which is defined as T _p = T _c • N, where
N is the number of packet stations for a given network;
The ability to achieve this goal, namely, to increase the efficiency of using the bandwidth of a shared channel with low traffic, is due to the fact that
- with this method of packet exchange time segments (T _c ) are relatively much shorter than in the prototype and they are used to transmit information and to determine the occupancy of the collective use channel;
- in the absence of information for transmission at a packet station, the segments allocated for this station can be used to transmit information from other packet stations;
- Each packet station has the ability to transmit its information in the allotted time for it.

 In FIG. 1 shows a structural electrical diagram of a packet station for implementing the proposed method; in FIG. 2 is a timing diagram of dividing a shared network channel of a local network into time segments, allocating segments to subscriber stations of a local network, forming packets when transmitting information from one, two and "N" transmitting stations; in FIG. 3 - graphs of the efficiency of using the bandwidth of the channel of collective use for the prototype and the proposed method.

 The device (Fig. 1) contains a timing unit 1, an access control unit 2, a receiver 3, a microprocessor 4, a transmitter 5.

The proposed method can be implemented as follows. Subscriber stations of the local network are constantly in a state of receiving information from a shared channel. From the moment the subscriber station is connected to the local network with an accurate time signal (for example, from the SEB) or another signal common to the local network, the timing unit 1 is started, by means of which time intervals T _c , t _{p are formed.} and t _op. (Fig. 2a, b, 2.1). Intervals T _c and t _op. allocated for the i-th subscriber station (where i = 1, 2, ..., N) in the control unit 2, to the other input of which from the receiver 3 there is a signal about the presence of a carrier in the shared channel (channel busy signal), the interval in "own" segment T _c , on which the carrier is analyzed (Fig. 2, b). If on the “own” interval T _{c the} absence of information in the communication channel is detected for t _op. , at the end of the interval T _c , an access permission signal is generated, which is transmitted to microprocessor 4 and determines the possibility of transmitting an information packet from microprocessor 4 through transmitter 5 to a shared communication channel. In this case, if there is information for transmission, the transmitter 5 is turned on for transmission and the information is transmitted from the microprocessor 4 in packets of the same length in a synchronous manner so that the beginning of the packet coincides with the beginning of the next segment beyond its own, and the transmission of the packet completely ends within the allotted time interval the value of which is defined as T _p = T _c N, where
N is the number of packet stations for a given local network, and T _c ≥ t _p. + t _op. where
t _r - the propagation time of signals between the two most remote packet stations in the local network;
t _op. - the carrier recognition time from other stations in the shared channel adopted in this local network.

 In FIG. 2.2 is a timing diagram of the use of a shared channel when transmitting packet information from a 2nd subscriber station, FIG. 2.3 is a timing chart for the case of transmission from the 1st and 3rd subscriber stations, in FIG. 2.N - from all N stations.

 Implementation of the proposed exchange method in a packet network allows to increase the efficiency of using the bandwidth of a collective use channel with low traffic (with the number of operating subscriber stations from 1 to N-1) by providing information transfer in the segments of those subscriber stations that do not currently transmit information and uniform distribution of segments of the shared channel between subscriber stations having information for transmission. In FIG. Figure 3 graphically shows the possibility of increasing the efficiency of using the bandwidth of a shared channel (C) depending on the number of packet stations (n) transmitting information.

 Sources of information.

 1. "Data transmission systems and computer networks", Moscow, Mir, 1974, p. 60.

 2. Scherbo V. K., Kirichev V. M., Samoilenko S. I. "Standards on local area networks", M., "Radio and communications", 1990.

 3. TIIER, M., Mir, Volume 75, N 1, 1987, p. 167.

Claims (2)

1. A method of packet exchange in local networks, in which all subscriber stations collectively use a common physical environment, subscriber stations synchronize with each other, divide the current time into segments and constantly receive information, characterized in that the current time is divided into segments of equal duration by T _s > = t _p + t _op, where t _p - propagation time of signals between the two most distant subscriber stations in the local network and the _op t - time identification carrier presence from other stations in the channel collectively of use, received in the LAN, provided undetected for a subscriber station within its interval T _to transmit information from the other network and the availability of information to send, it transmits the packet subscriber stations of the same length in a synchronous manner so that the start of the packet coincides with the beginning of the subsequent for "their" segment, and the transmission of the packet completely ended within the allotted time interval, the value of which is determined as T _p = T _with N, where N is the number of subscriber stations in this network. 2. The method of packet exchange according to claim 1, characterized in that the value of the packet is defined as T _p = T _with Nk, where k = 1, 2, 3, ....

Images