Printer Friendly

Contention Control Critical in Mobile Radio Datacomm System.

A mobile radio data communications system is more than just a voice communications system on which data is transmitted. The performance of data communications system is often compromised by the limitations of a radio system originally designed for voice transmission. The channel throughput for a typical system architecture is characterized for Aloha, slotted Aloha, carrier-sense multiple-access (CSMA) and digital-sense multiple-access (DSMA) contention techniques, with these techniques compared for efficiency in handling short and long message packets.

Conceptually, mobile radio data communications can be thought of as an extension of a packet data network into the mobile environment. Data packets transmitted on the radio are similar to X.25 Datagrams or X.25 packets for which a virtual circuit has already been established. The key to achieving high system throughput is to minimize the number of relatively short-system control packets required for each data transmission and judicious selection of the channel contention technique.

A typical layout of radio cells and frequency use within a cell is shown in Figure 1. A very small system may consist of one duplex radio channel used at one cell site, while a large system may involve a large number of cell sites with many radio channels at each cell site.

In Figure 2 a block diagram of a large system is shown. The public X.25 network is used to switch between subscriber host-computer equipment and base-station equipment located at the cell sites. Each radio channel at each cell site contains a communication controller, a base-station controller and a base-station radio transceiver.

Cell Site Elements

The communications controller interfaces the X.25 network, formats messages destined for the radio channel into a robust error-protected format, detects and corrects erros in messages received from the radio channel, and buffers and acknowledges messages in both directions. The base-station controller functions as a 4800 b/s radio modem and also generates the busy bit in the DSMA channel-contention technique. A conventional base-radio transceiver is used in this system.

In the system shown, a small subscriber may have one host computer in an office connected to 20 or 30 mobile-data terminals in different radio cells, through the X.25 public network and the base-station equipment shown. A mobile terminal initiates communications by providing the communications controller on its radio channel with the X.25 address of its host computer. The communications controller then establishes a switched virtual circuit (SVC) to the subscriber host computer. Data communications between host computer and mobile-data terminal can then proceed. The SVC is cleared when the mobile moves to a new cell or frequency, or after a period of inactivity.

Access to a shared channel results in contention for a limited resource. The channel utilization is defined by the level of successful message transmissions relative to the maximum channel capacity.

The simplest scheme for accessing a channel uses no contention control, allowing each mobile to transmit independent of the other mobiles and the state of the channel.

In a slotted Aloha contention control scheme, message transmissions are restricted to begin at the beginning of a fixed time slot. Thus, if two or more packets collide, they overlap completely, and only the given time slot is lost.

A much more efficient contention-control scheme is based upon preventing additional mobiles from transmitting when the channel is in use. The base system detects when a message packet is being transmitted and signals to the mobiles that the channel is busy. The usual method for sensing the busy condition of a channel is to detect the presence of the carrier signal. This scheme is the CSMA.

The performance of CSMA is dependent upon the ability of a mobile to determine the channel status with a minimum of uncertainty. The presence of a message transmission on the channel must be detected and the mobiles signalled with a minimal delay, for during this period additional transmissions may be initiated that will produce collisions. This delay is referred to as the collision interval and is the prime parameter in determining the channel utilization for CSMA.

The Mobile Data International (MDI) system uses a contention control technique based upon detection of digital data in a message transmission. DSMA is analogous to CSMA in its operation, but the use of digital data detection provides much better performance by reducing the channel busy detection delay.

The performance of the DSMA contention-control scheme is also improved by the use of a digital bit sequence to indicate the channel busy condition to the mobiles. The setting and detection of the digital busy bit is a simple and fast signaling technique that greatly reduces the channel signaling delay.

For a two-frequency full-duplex system where the base station is receiving continuously, the total collision interval is made up of the mobile transmitter attack time; propagation delay, mobile to base; time for base station to integrate data signal to obtain confidence that data is present rather than noise or voice; time for base station to set busy signal; propagation delay, base to mobile and inhibit transmit function.

The base station busy detect time for DSMA is equivalent to the data presence detection time, which depends on the modem design. For a 4800 b/s modem this can be as fast as 1 ms, but a longer period (10 ms) is typically required to reduce the probability of false busy detection (on noise bursts) to an acceptable level.

With no signals on the inbound channel, the detection time is normally adjusted to the shortest value at which there are no false busy indications. The setting will depend on the nature of the background noise and interference, which in turn depends on the RF environment at the site and the characteristics of the type of radios used.

The impact of false busy detection on message delay and channel throughput can be very significant. Every time a false busy is detected, channel capacity is wasted because no terminal will transmit. Even a slight decrease in channel capacity can result in a large increase in the average delay per message.

If mobiles are permitted to transmit carrier without data (voice only), then the base busy detect time must be increased significantly and the presence of carrier without data can also be used in determining if the channel is busy.

The base busy time and mobile busy detect time is the time taken for the base station to set the busy signal plus the time taken by the mobile to detect it. The MDI system uses a digital method to set and detect channel busy. The base station continuously transmits a bit pattern that is changed when the receive channel is busy. When an inbound signal is detected at the base station receiver, the digital busy indicator is set.

Channel Utilization Reduced

The channel utilization is reduced by the delay in detecting that the channel is no longer busy. This delay, called the busy hang time, can be varied as is done for the data presence detection. The busy hang time must be set so that it is longer than channel fades, otherwise new transmission may collide with an ongoing one. However, the busy hang time must not be set too long, because it constitutes lost-channel capacity.

The performance relationship between maximum channel utilization and the collision interval is shown in Figure 3. For small collision intervals, the maximum throughput capacity for CSMA or DSMA contention control is far greater than the maximum capacities of Aloha and slotted Aloha.

The performance of non-persistent CSMA and DSMA contention control depends on the duration of the collision interval relative to the length of the transmitted message packets. In a typical system, the collision interval is kept as small as possible, and as an example may be in the order of 15 milliseconds. However, the length of the message packets can typically vary from 500 to 4000 bits, or 100 to 900 milliseconds in duration. Thus, the relative collision interval varies from 15 percent to 60 percent depending upon the message length.

The maximum channel utilization for a fixed-collision interval depends, therefore, on the length of the message packets. The longer the data packet, the greater the data throughput attainable.

Non-persistent CSMA and DSMA and highly efficient contention techniques provided that the collision interval and the busy hang time can be minimized. In addition, best results are obtained when the message packet length is relatively long compared with the collision interval.
COPYRIGHT 1985 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1985 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Morris, J.; Kenward, G.
Publication:Communications News
Article Type:evaluation
Date:May 1, 1985
Previous Article:Increased Competition Broadens International Services Menu.
Next Article:Technical Advances in Data Communications Alter Large-Scale Messaging Environment.

Related Articles
Fourth-Generation Protocol Analyzers Support Data Communications Managers.
Procomm Alaska: Motorola communications systems for Alaska from offices in Anchorage, Fairbanks, Wasilla, Kenai. (Business Profile: Advertisement).

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters