WO2002043318A2

WO2002043318A2 - Media access control with echo based collision detection and resolution

Info

Publication number: WO2002043318A2
Application number: PCT/IL2001/000921
Authority: WO
Inventors: Mordechai Mushkin; Dan Raphaeli
Original assignee: Itran Communications Ltd.
Priority date: 2000-11-22
Filing date: 2001-10-03
Publication date: 2002-05-30
Also published as: WO2002043318A3; AU2001295861A1

Abstract

A novel and useful apparatus for and method of echo based collision detection and resolution used in controlling access to a shared media. The mechanism utilizes echoes generated by non-competing nodes to enable operation in the presence of hidden nodes. Mechanisms are provided for use by source and destination nodes to compete for the media. In addition, two mechanisms for handling the blocked node problem are provided. The first is a mechanism for initiating a link by a potential destination node, rather than the source node. The second is a mechanism whereby a node transmits a release indication signal to the media upon detecting that the media is available after being busy for a relatively long period of time. The invention also includes a mechanism for multi-priority level occupation of the media whereby a node may capture the media at different priority levels.

Description

MEDIA ACCESS CONTROL WITH ECHO BASED COLLISION DETECTION AND RESOLUTION

FIELD OF THE INVENTION The present invention relates generally to data communication systems and more particularly relates to a media access control method utilizing echo based collision detection and resolution.

BACKGROUND OF THE INVENTION Carrier Sense Multiple Access (CSMA) is a well-known media access mechanism, which is used for example in Ethernet LANs (IEEE Standard 802.3) and wireless LANs (IEEE Standard 802.11). They are commonly used in networks that comprise shared media whereby multiple nodes simultaneously have access to the same media. The media may be any physical medium that can be simultaneously shared by many nodes, such as a cable, RF, powerline, etc.

An example of a shared network is shown in Figure 1A. Referring to Figure 1A, the network, generally referenced 10, comprises a plurality of communication nodes 12 (nodes A through E) that are connected to a common physical media. Each node 12 is connected to a shared media 14. An example of a shared media is the AC powerline wiring grid found in homes, offices and factories. In a residential environment, groups of neighboring residences are coupled together via the outdoor wiring, thus forming a huge common media. Note that the powerlme media remains shared until reaching a transformer where signals cannot easily propagate beyond without signal couplers.

A shared media such as the powerline is typically characterized by a large variety of different signal propagation conditions. In many cases, portions of the media are invisible from other parts creating hidden node situations. In the example shared network shown in Figure 1 and those presented hereinbelow, it is assumed that nodes that appear adjacent in the drawing can hear each other but non adjacent nodes cannot.

In many cases, a home, enterprise or other premise includes more than one communication network. Each communication network may be made up of a plurality of nodes with each network comprising at least two nodes. All nodes of the same network implement the same communication technique and are able to communication with each other thus permitting interoperability (assuming that the propagation conditions over the media enable communication). Nodes from different networks may implement different communications techniques, in which case they are not able to communicate with each other. In addition, the propagation characteristics of the shared media (e.g., the powerline grid) may have large variations and irregularities. This results in large variations in the attenuation over the communication path between two given nodes. Since the powerline grids of neighboring residences are physically connected via the power distribution network, the common media of Figure 1A might refer to the powerline grid of a single residence or to the powerline grids of several neighboring residences (e.g. several apartments in a building).

In a configuration such as in Figure 1A, reliable efficient communications between the nodes is usually possible only if one node transmits at a time, in order that the transmission from one node does not interfere with the transmission from another node. Therefore, there is a need for a media access mechanism that enables nodes to share the media without interfering with each other. This applies to nodes of the same network as well as to nodes of different networks. Thus, it is desirable to have nodes that belong to different networks but share a common media be able to coexist with each other. Coexistence, i.e. media sharing, entails nodes from one network recognizing when nodes from another network desire to communicate and refraining from initiating new transmissions until the ongoing transmission is complete. However, nodes from different networks typically utilize different communication protocol stacks, thus preventing them from detecting and understanding each other's messages (typically within the Physical layer).

In a network incorporating prior art CSMA like media sharing, a node that desires to transmit, captures the media by the mere transmission of the data message or by transmitting before or along with the data message a predefined signal or message. As long as the media is captured, no other node is permitted to transmit.

In basic CSMA, a node that desires to transmit listens to the media to determine whether any other node is transmitting and transmits only when the media is free. A variety of prior art carrier sense mechanisms have been developed including physical carrier sense and virtual carrier sense. The principle of physical carrier sense is the direct detection of carrier signal energy. The actual transmission is continuously detected as long as it exists, hence the derivation of the term 'carrier sense.' This mechanism is incorporated in the IEEE 802.3 Ethernet LAN standard. It is suitable for applications where every node is capable of determining whether any other node is transmitting.

This carrier sense mechanism, however, has the disadvantage of poor reliability in noisy channels where real signal cannot always be discerned from noise. In addition, it does not permit the coexistence of nodes having different technologies. Further, the mechanism is susceptible to the hidden node problem.

The previous invention described in detail in co-pending U.S. Application Serial

No. 09/679,330, entitled "entitled "Media Access Control Utilizing Synchronization

Signaling," referenced above, provided a novel media access control (MAC) mechanism that utilizes synchronization signaling to enable nodes from different networks having different technologies and protocols to coexist using the same shared media.

The media access control mechanism of the previous invention provides a solution to the communication problems imposed by the characteristics of the powerline media, e.g., the hidden node and mask node problems. In addition, the MAC mechanism enables the efficient coexistence of nodes that employ different communication technologies allowing multiple networks attached to the shared media to coexist regardless of the particular technology they use.

The MAC mechanism comprised assigning specific time slots for the transmission of a busy signal (or frame occupation signal), which functions to indicate to all nodes that hear it that the media channel is occupied. The frame occupation signaling is transmitted periodically at the same time during each frame. A node refrains from initiating a new transmission if it detects the frame occupation signal. A node begins transmission only after it ceases to detect a frame occupation signal. This provides a method to ensure that when a given group of nodes (typically a pair of nodes) have successfully captured the media, all other nodes will refrain from initiating new transmissions as long as the group of nodes occupies the media.

The approach to the problem of collisions, however, is handled by the previous invention using well-known collision avoidance techniques. The term 'collision' refers to the situation when two or more nodes begin a new transmission over an available media at the same time. Within the framework of the previous invention, a collision occurs when more than one node begins as transmission within the same available frame. The result of a collision might be that the messages transmitted by the colliding nodes will interfere with each other and both destination nodes will be prevented from properly receiving the messages. Therefore, if collisions are not handled properly, it is very likely that utilization of the media will be very poor or that communications may not be possible altogether.

The problem presented by the occurrence of collisions is a problem common to all CSMA and CSMA like mechanisms. Conventional methods to handle collisions are known and include collision avoidance, collision detection and collision resolution. Each is described hereinbelow.

If all nodes that wish to capture the media do it at the first available point in time point, i.e. at the first available frame assuming the framework of the MAC mechanism described in the previous invention, the probability of collision will be very high. Further, collisions may continue indefinitely if, after a failure due to a collision, the competing nodes re-attempt to capture the media at the next available frame.

Collision avoidance is a well-known method that is effective in reducing the probability of collisions. Collision avoidance methods are used in many CSMA media access protocols, for example in IEEE 802.11. In this method, a node that wishes to capture the media does not do it immediately when the media becomes available. Rather, the node waits a random period of time. This waiting is usually referred to as 'random back off.' In the shared media framework described in the previous invention, collision avoidance means that rather than a source transmitting at the first available frame, the source waits a random number of frames until it attempts to capture the media. A major disadvantage of collision avoidance is that although it is effective to reduce the probability of collision, the penalty is time wasted performing the random back off.

In collision detection, each competing node transmits a special signal that is intended to be unique to the node at that moment (e.g., a randomly chosen signal). While the competing node transmits the collision detection signal, it simultaneously listens to the media. If a given competing node detects a signal that is different from the collision detection signal it transmitted, the node given assumes that another node is competing for the media and the given node performs a random back off.

Shown in Figure 1A is a generic example of several nodes sharing the same media. Note that the same generic example is used in a number of places hereinbelow to demonstrate several network situations. The nodes in the generic example are labeled A, B, C, D and E. It is assumed that the path attenuation conditions across the network are such that 'neighboring' nodes (i.e. A and B, B and C, C and D and D and E) are able to detect each other's transmissions while and non-neighboring nodes cannot detect each other's transmissions. It is also assumed that simultaneous transmissions from both neighbors of a given node interfere with each other but transmissions from non- neighboring nodes do not interfere. For example, simultaneous transmissions from nodes A and C interfere at node B such that node B is not aware of the transmissions. Note that the terms 'adjacent' and 'neighboring' as used herein refer solely to the alphabetic order of the letters indicating the nodes (i.e. A, B, C, D and E) and to the location of the nodes depicted therein. It does not necessarily have any relationship with the physical location of the actual nodes.

Based on the generic network example of Figure 1A, Figure IB presents an example of collision detection. Assume node B wishes to transmit to node A and node C wishes to transmit to node D. Assume also that nodes B and C transmit different collision detection sequences. Nodes B and C detect each other's collision detection sequence and both refrain from capturing the media.

The collision detection sequences are shown in detail in Figure lC. Node B transmits sequence 16 while node C transmits sequence 18. The hatched boxes or bins represent the presence of signal (i.e. a '1') while empty bins represent the absence of signal. Node B detects sequence 20 while node C detects sequence 22. The sequences detected by both nodes do not match their originally transmitted sequences, thus they both back off from the media. Collision detection is an improvement in efficiency over collision avoidance because with collision detection, back off is employed only after a collision has been detected. A disadvantage, however, is that collision detection fails in the presence of hidden node, as illustrated below.

The hidden node problem is described as follows with reference to Figure ID. Assume node A is transmitting to node B. Node C is unaware of the transmissions due to the propagation conditions of the example, i.e. it is 'hidden' from node A. Meanwhile, node C begins transmitting to node D resulting in a disturbance in reception at node B. Figure IE presents an example shared network designed to illustrate the effect of hidden nodes on the process of collision detection. In this example, both node B and node D wish to capture the media and communicate with node C. Node B and node D transmit different collision detection signals. Since nodes B and D do not hear each other, each detects the same signal that it transmitted without any change. Thus, both node B and node D assume that there is no other competing node and both capture the media. However, since the messages from node B and node D interfere with each other at the receiver of node C, none of the messages are detected and a link is not established.

Conventional collision resolution is similar to collision detection in the sense that each competing node transmits a special signal unique to the node. The competing node transmits the collision resolution signal while simultaneously listening to the media. If a given competing node detects a signal that is different from the collision resolution signal it has transmitted, the given node concludes that another node is also competing for the media and immediately stops transmission.

The collision resolution sequences are shown in detail in Figure IF. Node B transmits sequence 24 while node C transmits sequence 26. The hatched boxes or bins represent the presence of signal (i.e. a '1') while empty bins represent the absence of signal. Node B detects sequence 28 while node C detects sequence 30. In operation, node C detects node B's transmission and stops capturing the media. Note B detects no transmissions from other nodes and thus survives to capture the media. Collision resolution improves the collision handling efficiency by eliminating the time wasted during back off. Whenever a node wishes to capture the media, the collision resolution process is performed at the first available frame with a high probability that only a single surviving node will successfully capture the media.

A disadvantage, however, is that collision resolution fails in the presence of hidden nodes similarly to collision detection. Note that the example given in Figure IE is connection with collision detection is applicable to the case of collision resolution as well.

In summary, collision avoidance is effective to reduce the probability of collision, but suffers from a significant loss in efficiency due to the time wasted in performing the random back off. Collision detection and especially collision resolution are effective in improving the efficiency by reducing or even eliminating the time wasted on random back off. Both collision detection and collision resolution, however, do not operate properly in the presence of hidden nodes, a situation common to many types of media.

Therefore, there is a need for a collision detection and resolution mechanism that is capable of detecting and resolving collisions while operating in the presence of hidden nodes. SUMMARY OF THE INVENTION Accordingly, the present invention provides a novel and useful apparatus for and method of echo based collision detection and resolution. The echo based collision detection and resolution mechanism is used in controlling access to a shared media. The mechanism utilizes echoes to allow operation in the presence of hidden nodes. The method of the present invention is particularly useful in communication systems characterized by shared media such as networks that use the powerline as the media.

In accordance with the invention, in echo based collision detection, competing nodes attempt to capture the media by transmitting a unique collision detection or collision resolution sequence. Non-competing nodes that hear the sequence echo it back onto the media. The competing nodes hear the echo and compare it to the originally transmitted sequence. If they match, the node assumes the media is available and proceeds to capture it and begin transmitting. If they do not match, however, the node assumes that at least one other competing node is also trying to capture the media and as a result, the node backs off. In echo based collision detection, the original sequence and the echo sequence are allocated separate fixed locations (i.e. time slots) within the frame. The echo sequence located at a position later in time within the frame.

Echo based collision resolution is similar to detection with the difference being that the original sequence and the echo sequence are interleaved together bin by bin. The competing node transmits the original sequence one bin at a time. The non-competing nodes that hear the sequence bin, echo what they hear back onto the media. The competing node listens to the echo sequence bin and compares it to the sequence bin originally transmitted. If they match, the node continues transmitting the sequence. If it does not match, the node ceases transmission immediately and refrains from competing for the media in this frame.

If each competing node uses a unique collision resolution sequence, it is guaranteed that a single node with ultimately survive. The node to survive is the node with the collision resolution sequence having the largest binary value.

A key advantage of the echo based collision detection and echo based collision resolution methods of the present invention is that they enable the capture of the media by competing nodes in the presence of hidden nodes. The problem of hidden nodes is overcome by the non-competing nodes echoing the collision sequence they hear back onto the media.

In addition, the present invention provides a mechanism for destination nodes to compete for the media in addition to providing a mechanism to capture the media for source nodes. The invention also provides two mechanisms for handling the blocked node problem. The first is a mechanism for initiating a link by a potential destination node, rather than the source node. The second is a mechanism whereby a node transmits a release indication signal to the media upon detecting that the media is available after being busy for a relatively long period of time. The invention also provides a mechanism for multi-priority level occupation of the media whereby a node may capture the media at different priority levels. An interrupt and interrupt echo mechanism is provided to enable a node at a higher priority to interrupt the communications of a lower priority node.

The invention also provides a power level regulation mechanism whereby the transmission power level and the detection threshold of the busy signals is refined, regulated and adjusted to the power level of the potential interfering transmitter and to the acceptable interference level at the potentially interfered receiver.

Many aspects of the previously described invention may be constructed as software objects that execute in embedded devices as firmware, software objects that execute as part of a software application on a computer system running an operating system such as Windows, UNLX, LINUX, etc., an Application Specific Integrated Circuit (ASIC) or functionally equivalent discrete hardware components.

There is thus provided in accordance with the present invention a method for detecting collisions among a plurality of competing source nodes in a distributed network of nodes connected to a media, the method comprising the steps of each competing node transmitting a collision detection sequence onto the media, each non-competing node monitoring the media and listening for a collision detection sequence, each non-competing node echoing the collision detection sequence heard thereby, each competing node listening for the echoed collision detection sequence and comparing its originally transmitted collision detection sequence therewith, backing off from transmission if the echoed collision detection sequence does not match the collision detection sequence previously transmitted and transmitting a message onto the media if the echoed collision detection sequence does match the collision detection sequence previously transmitted. There is also provided in accordance with the present invention a method for resolving collisions among a plurality of competing source nodes in a distributed network of nodes connected to a media, the method comprising the steps of each competing node transmitting a collision detection sequence one bin at a time onto the media, each non- competing node monitoring the media and listening for a collision detection sequence bin, each non-competing node echoing the collision detection sequence bin heard thereby interleaved with the transmission of the collision detection sequence, each competing node listening for the echoed collision detection sequence bin and comparing the collision detection sequence bin previously transmitted thereby, backing off from transmission if the echoed collision detection sequence bin does not match the collision detection sequence bin previously transmitted and transmitting a message onto the media if the echoed collision detection sequence bin does match the collision detection sequence bin previously transmitted.

There is further provided in accordance with the present invention a method for initiating a link between a potential source node and a blocked destination node in a distributed network of nodes connected to a media, the method comprising the steps of determining that a node is a potential blocked destination node, the blocked destination node attempting to capture the media and once the media is captured, establishing a connection between the destination node and the source node. There is also provided in accordance with the present invention a method for indicating the release of a media by a node in a distributed network of nodes connected to the media, the method comprising the steps of a node detecting that the period of time the media has been busy exceeds a threshold, detecting that the media has become available, transmitting a release indication signal onto the media indicating that the media is now available and a potential source node detecting the release indication signal and in response thereto attempting to capture the media.

There is still further provided in accordance with the present invention a method of prioritized media capture for use in a distributed network of nodes connected to a media, the method comprising the steps of transmitting a low priority busy signal when a node occupies the media at low priority, transmitting a high priority busy signal when a node occupies the media at high priority, transmitting an interrupt signal when a node wants to compete for the media at high priority, each potential high priority destination monitoring the media for the presence of an interrupt signal; and if detected, echoing the interrupt signal as an interrupt echo signal, each node occupying the media at low priority monitoring the media for the presence of either the interrupt signal or the interrupt echo signal; and in response to the detection thereof and releasing the media.

There is also provided in accordance with the present invention a method for dynamically regulating the power level of busy signals transmitted by nodes in a distributed network of nodes connected to a media, the method comprising the steps of adjusting the transmission power of the busy signals transmitted by a node to a level such that the busy signals are detected by nodes potentially interfering with the node while they are not detected by nodes sufficiently distant from the node to be non-potentially interfering nodes and adjusting the detection threshold of busy signals received by the node such that the busy signals transmitted by potentially interfering nodes are detected by the node while the busy signals transmitted by non-potentially interfering nodes are not detected by the node.

BRIEF DESCRIPTION OF THE DRAWINGS The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein:

Figs. 1A, IB, ID and IE are diagrams illustrating examples of shared networks; Figs. 1C and IF are diagrams illustrating examples of collision detection and resolution;

Fig. 2 is a diagram illustrating the collision detection sequence and echo sequence fields inserted into the frame in accordance with the present invention;

Fig. 3 is a flow diagram illustrating the echo based collision detection method of the present invention;

Fig. 4 is a diagram illustrating an example of the echo based collision detection method of the present invention;

Fig. 5 is a diagram illustrating the collision resolution sequence interleaved with the echo sequence fields and inserted into the frame in accordance with the present invention; Fig. 6 is a flow diagram illustrating the echo based collision resolution method of the present invention;

Fig. 7 is a diagram illustrating an example of the echo based collision resolution method of the present invention;

Fig. 8 is a diagram illustrating interference due to transmissions by a destination node;

Fig. 9 is a diagram illustrating the source resolution and destination resolution fields inserted into the frame in accordance with the present invention;

Fig. 10 is a flow diagram illustrating the echo based collision resolution method of the present invention incorporating competition by destination nodes; Fig. 11 is a diagram of a shared network illustrating the blocked node problem;

Fig. 12 is a flow diagram illustrating the link initiation by a potential destination node method of the present invention;

Fig. 13 is a diagram illustrating the release indication, collision resolution sequence and echo fields inserted into the frame in accordance with the present invention; Fig. 14 is a flow diagram illustrating the release indication method of the present invention; Fig. 15 is a diagram illustrating the high priority busy, interrupt, interrupt echo and low priority busy fields inserted into the frame in accordance with the present invention;

Fig. 16 is a flow diagram illustrating the multiple priority communication interruption method of the present invention; Fig. 17 is a diagram of a shared network illustrating the range of the busy signal transmitted by a node;

Fig. 18 is a flow diagram illustrating the power level regulation method of the present invention; and

Fig. 19 is a block diagram illustrating a node device incorporating a media access controller constructed in accordance with the present invention.

DETAILED DESCRIPTION OF THE INVENTION

Notation Used Throughout The following notation is used throughout this document.

Term Definition

AC Alternating Current

ASIC Application Specific Integrated Circuit

CSMA Carrier Sense Multiple Access

EEROM Electrically Erasable Read Only Memory

FPGA Field Programmable Gate Array

IEEE Institute of Electrical and Electronic Engineers

ΓR Infrared

LAN Local Area Network

MAC Media Access Control

OSI Open Systems Interconnect

RAM Random Access Memory

RF Radio Frequency

ROM Read Only Memory

Detailed Description of the Invention

The present invention is an echo based collision detection and resolution mechanism used in controlling access to a shared media. The mechanism utilizes echoes to overcome the problems posed by the presence of hidden nodes. The method of the present invention is particularly useful in communication systems characterized by shared media such as networks that use the powerline as the media.

For purposes of this specification, the term 'node' shall be taken to mean any network entity, implemented in either hardware or software, which may be the endpoint of a call or link within a shared network. The network may comprise any type of shared network or media including but not limited to power line based networks, twisted pair networks, IR wireless networks, RF wireless networks, optical fiber ring networks, etc. The term 'call' or 'link' shall be taken to mean any communication path that is established between at least two nodes for the purpose of communication therebetween.

Echo Based Collision Detection Mechanism The echo based collision detection mechanism of the present invention will now be described in more detail. The novel method of echo based collision detection, described hereinbelow, improves upon the conventional collision detection technique. The purpose of the invention is to provide a collision detection process that is capable of operating in the presence of hidden nodes.

In this echo based collision detection mechanism the competing node transmits a collision detection sequence at a predefined location in the frame. In other words, the collision detection sequence is transmitted at a predefined time slot in the frame. Assume that communications over the shared media occur by passing frames between nodes. During the transmission the competing node also monitors the media for the possible detection of another competing node.

A diagram illustrating the frame structure including collision detection sequence and echo sequence fields inserted into the frame in accordance with the present invention is shown in Figure 2. The frame sequence, generally referenced 40, comprises a sequential stream of frames 42, labeled frame N-l, frame N and frame N+l. The contents of a free frame include, inter alia, a busy field 41, synch sequence field 43, a collision detection sequence 44 and an echo collision detection sequence 46. A flow diagram illustrating the echo based collision detection method of the present invention is shown in Figure 3. A key feature of this method is that every non-competing node listens for and detects the collision detection sequence and repeats it in a predefined location in the frame (i.e. a predefined time slot within the frame).

Initially, each node that desires to capture the media, transmits a collision detection sequence (step 60). Each node that desires to capture the media is termed a competing node. Non-competing nodes may fall into one of two groups: those that are capable of being destinations and those that are not (a node that detects a busy signal at the frame header is not able to be a destination in this frame). The signal used for collision detection should allow simultaneous transmission and detection and should allow detection in the presence of different signals. This is usually achieved by an 'on/off keying type of sequence. The on/off sequence comprises a sequence of bins (i.e. a location in time) and each bin might be in one of two states: (1) 'on' which is indicated by a pulse being transmitted, or (2) 'off which is indicated by the absence of a transmission. The value of the on/off sequence is usually selected randomly but may be fixed for a particular node. Each competing node then monitors the media during the off bins and cases transmission if a collision is detected (step 62). Collision detection is performed as follows by each competing node transmitting an on/off sequence and listening to the media during off bins. If a signal is detected during one of the off bins, the existence of another competing node is concluded. Note that it is preferable that the pulses used for collision detection have the collision resistance feature described in the previous invention. This feature enables several pulses arriving simultaneously from different nodes not to cancel each other whereby the combined signal is a detectable pulse. Non-competing nodes listen to the media for a collision detection sequence (step

64). Note that only the non-competing nodes that are capable of being destination nodes need to listen, because they are the only nodes which are relevant for collision detection, and also because that they are the only nodes that are permitted to echo (a node that detects a busy signal in the frame header is not allowed to transmit in that frame). The non- competing nodes echo the sequence they receive from the media (step 66). Competing nodes then detect the echoed sequence (step 68) and compare it with the sequence they originally transmitted (step 70).

If the echoed sequence is different from the transmitted one, the competing node concludes than there exist at least one other competing node and the node backs off (step 78). Otherwise, the competing node assumes that the media is available since it is the only competitor for the media (step 74) and the node transmits its message to its destination

(step 76).

A diagram illustrating an example of the operation of the echo based collision detection method of the present invention is shown in Figure 4. The situation in this example are the same as in Figure IE, wherein both node B and node D wish to communicate with node C and both attempt to capture the media on the same frame. Note that in accordance with the method, node B and node D transmit different random sequences. In this example, the sequence chosen by node B is 101101 and the sequence chosen by node D is 110001. In accordance with the method of the present invention, node A receives the sequence transmitted by node B and in response thereto echoes it back onto the media.

Similarly, node E receives the sequence transmitted by node D and echoes it. Node C receives the sequences from both nodes B and bode D. The sequence actually received by node C is the ORing of the two original sequences which results in 111101. This is the echo sequence node C echoes back into the media.

Node B simultaneously receives the echoes transmitted by nodes A and C, and detects the sequence 111101, which is different from node B's original sequence 101101. Therefore, node B assumes there is another node competing for the media and it backs off. Similarly, node D detects the ORing of the echo sequences transmitted by nodes C and E which is different that its originally transmitted sequence and it also backs off from the media.

It is important to note that the example presented above illustrates that the echo based collision detection method of the present invention operates properly in the presence of hidden nodes. Nodes B and D are hidden to each other, but the echo mechanism enables them to be noticed by each other during the collision detection phase of the media capture process.

Echo Based Collision Resolution Mechanism The echo based collision resolution mechanism of the present invention will now be described in more detail. The echo based collision resolution method of the present invention improves upon the conventional collision resolution technique. The purpose of the invention is to ensure proper operation of the collision resolution mechanism especially in the presence of hidden nodes. The basic operation of this mechanism is similar to that of the echo based collision detection described hereinabove, with the appropriate modifications to provide collision resolution. In this method, the competing node transmits a collision resolution sequence at a predefined location within the frame. During the transmission the competing node also monitors the media to detect the possible presence of one or more competing nodes. Each non-competing node, which is able to be a destination, is operative to detect the collision resolution sequence and repeat it in a predefined location in the time frame. However, unlike in echo-based collision detection, in echo based collision resolution, the echoed sequence is interleaved with the transmitted sequence as shown in Figure 5. Each of the frames 82 in the frame stream, generally referenced 80, comprise, inter alia, a busy field 81, sync sequence field 83 and a combined sequence and echo field 84. This field comprises bins of the transmitted collision resolution sequence 86, labeled Si, S₂, S₃, ..., SM. interleaved with bins of the echoed collision resolution sequence 88, labeled Ei, E₂, E₃, .., E_M- Note that every original sequence bin is followed by a respective echoed sequence bin. A flow diagram illustrating the echo based collision resolution method of the present invention is shown in Figure 6. First, all competing nodes that desire to capture the media transmit a collision sequence bin by bin (step 100). The signal used for collision resolution must ensure that only a single node will survive if all competing nodes use (either randomly chosen or pre-assigned) different collision resolution signals. The signal used for collision resolution can be similar to the signal used for collision detection, i.e. an on/off keying sequence that has a collision resistance property. Each competing node monitors the media during the off bins and ceases transmission is a collision is detected (step 102). Collision detection is performed by each competing node transmitting an on/off sequence bin and listening to the media during the off bins. If a signal is detected during an off bin, the existence of another competing node is concluded, and the node immediately stops transmission. Note that whenever several nodes with different sequences compete for the media only one node will survive. If the random sequences are taken to represent binary numbers, and if an 'on' bin represents a binary '1' and an 'off bin represents a binary '0,' it can be easily shown that the node with the highest number will be the surviving node.

The non-competing nodes that are allowed to transmit listen to the media for collision detection sequence bins (step 104). Those non-competing nodes echo the collision resolution sequence bin heard in an interleaved manner with the original sequence bins (step 106). Competing nodes listen for the echoed collision resolution sequence and compare it to the sequence bin originally transmitted (step 108). If a competing node detects an 'on' echo bin following an 'off sequence bin transmitted by itself (step 110), the node concludes that another node is also competing for the media and immediately stops transmission (step 116).

If there are additional bins to be transmitted (step 111), the method repeats with step 100. If after all the echo sequence bins are received no 'on' echo bin following an 'off sequence bin was detected, the competing node assumes that media is available (step 112) and begins transmission (step 114). Note that if each competing node transmits different random collision resolution sequences, only one node will survive, the node with the longest binary collision resolution sequence.

A diagram illustrating an example of the operation of echo based collision resolution method of the present invention is shown in Figure 7. The example is similar to that of Figure 4 described supra wherein both node B and node D wish to communicate with node C and both attempt to capture the media on the same frame. Note that in accordance with the method, node B and node D transmit different random sequences. In this example, the sequence to be transmitted by node B is 101101 and the sequence to be transmitted by node D is 110001.

In accordance with the method of the present invention, node A receives the first sequence bin (i.e. '1') transmitted by node B and in response thereto echoes it back onto the media. Similarly, node E receives the sequence bin ('1') transmitted by node D and echoes it. Node C receives the sequence bins from both nodes B and bode D. The echo sequence bin actually received by node C is the ORing of the two original sequence bins which results in '1'. This is the echo sequence bin node C echoes back into the media.

Node B simultaneously receives the echo bin transmitted by nodes A and C, and detects the echo sequence bin '1' which matches node B's original sequence bin '1'. the second sequence bin, however, node A receives the sequence bin '0' and node E receives the sequence bin '1'. Note C receives sequence the OR of sequence bins '0' and '1', i.e.

'1' which it echoes back to the media during the echo bin. Node B hears a '1' following its

'0' sequence bin and thus concludes that there is another node competing for the media and it stops transmitting. Node D, however, detects the ORing of the echo sequences transmitted by nodes C and E which matches its originally transmitted sequence and in response it captures the media and proceeds with transmission.

The difference being this collision resolution method and the previous collision detection method that in this method, node B detects an echo of Tat the second bin and stops transmission immediately. The result is that node D remains the only competitor and captures the media.

The above example illustrates that echo based collision resolution operates properly in the presence of hidden nodes. Note that nodes B and D are hidden to each other but the echo resolution mechanism of the present invention causes them to be noticed by each other during the collision resolution process.

Capture Mechanism Including Competition For Destination Nodes

The problem of interference due to destination nodes and a solution thereto will now be described in more detail. When a group of nodes capture the media and transfer data over it there is typically one node that initiates the capture. This node is referred to as the source node or the initiator. The other nodes (typically there is only one) are referred to as the destination node(s) or the responder(s). some cases the data transaction is unidirectional, such as from source to destination(s). In other cases the data transactions might be bidirectional. In either case, however, transmissions take place at both sides of the link because the destination nodes are usually expected to return information of some sort, such as acknowledgement messages, etc.

Therefore, when the media is captured, both source and destination nodes have to compete for it. An example of this need illustrating the interference due to transmissions by a destination node is shown in Figure 8. In this example network, generally referenced 120, node A wishes to communicate with node B and node D wishes to communicate with node C. Nodes A and D thus compete for the media, and both survive. Nodes A and D than begin transmitting to nodes B and C, respectively. Nodes B and C also respond. If it so happens that node B's response coincides with node's D transmission, interference will be caused at the receiver of node C. Thus, destination nodes must also compete for the media in order to avoid potential interference. The present invention provides a solution to this problem by inserting into the frame structure, in addition to the source resolution field described above, another field for destination resolution. A diagram illustrating the source resolution and destination resolution fields inserted into the frame in accordance with the present invention is shown in Figure 9. Each frame 132 in the frame stream, generally referenced 130, comprises, inter alia, a busy field 131, sync sequence 133, a dedicated place for source resolution (or competition) 134 is allocated near the beginning of the frame and a dedicated place for destination resolution (or competition) 138 is allocated toward the end of the frame. The source transmissions 136 follow subsequent to the source resolution and transmissions 140 in either direction follow subsequent to destination resolution 138. In accordance with the present invention, the source competition is performed using the echo-based collision resolution method described hereinabove. Destination competition, however, is performed using conventional collision resolution. Transmissions by a survivor source node may occur anywhere after the source collision resolution, excluding the destination collision resolution period. Transmissions by a survivor destination node may occur after the destination collision resolution period.

As described above, the echo based collision resolution mechanism, which is required for source competition, is not required for the destination resolution. The reason being that when the sources compete for the media, the destinations are not yet known. Therefore, when the sources compete all potential destinations participate by way of the echo mechanism. When the destinations compete, however, the situation is different. The destination(s) is (are) already known and it is already ensured that none of the destinations conflicts with any of the sources. Thus, the only competition that exists is among the destinations themselves, which can be easily resolved using conventional collision resolution without the need for the echo mechanism.

A capture mechanism including competition for destination nodes will now be described in more detail. A flow diagram illustrating the echo based collision resolution method of the present invention incorporating competition by destination nodes is shown in Figure 10. For illustration purposes, the method is described in the context of a source node A and a destination node B.

Initially, node A waits for the first available frame (step 150). Once an available frame is found, node A compete for the media using the echo based collision resolution mechanism (step 152). If any echoed bin was different than the sequence bin originally transmitted (step 154), node A stops transmitting and waits for the next available frame (step 164).

Otherwise, if all echoed bins match the sequence bin originally transmitted (step 154), node A sends a message to node B (step 156). In response, node B competes for the media with other potential destinations using conventional collision resolution (step 158). If node B fails to capture the media (step 160), it does not respond and node A does not proceed with the capture process (step 166). Note that node B may fail to respond if it is blocked or if other nodes are competing for the destination capture. If node B was able to respond and did survive (step 160), it responds to node A. From that point on nodes A and B can freely communicate with each other until the end of the frame, (step 162). They might also continue to communicate freely during the following frames, as long as they keep the media occupied by transmitting busy signals.

Blocked Node Handling Method

The blocked node situation handling method of the present invention will now be described in more detail. Note that the term 'blocked node problem' is used herein to refer to the situation wherein the media is available from the point of view of a given potential source yet it is busy from the point of view of the intended destination node. This problem is inherent to all CSMA and CSMA-like media access protocols. The present invention provides two methods for handling the blocked node situation within the context of a shared network: (1) a link initiation method whereby a potential destination node initiates the link rather than the source node and (2) and a release indication method whereby a node that has experienced a busy media condition for a long period of time is operative to transmit a signal indicating the recent release of the media. It is appreciated that the methods described hereinbelow may be used in many different frameworks (e.g., shared networks, etc.), with minor modifications possible required, and are not limited to the example frameworks described herein.

The blocked node problem will now be described in more detail. The blocked node situation is illustrated by the example shared network presented in Figure 11. In this example network, generally referenced 170, nodes C and D have captured the media and are keeping the media busy in order to maintain communications between each other, using for example, the busy signaling mechanism described in the previous application cited supra. Therefore, the media as seen by nodes B and E is busy. Assume now that node A wishes to communicate with node B. Node A competes for the media, survives and transmits a message to node B, but node B is unable to respond because the media as seen by node B is busy. Note that it may also be the case that node B does not receive node A's message because of interference by the transmissions of node C.

Node A, however, has no indication about the availability of the media at node B, and thus node A does not know the reason why node B fails to reply. If node A persists on trying to initiate communication with B at every available frame, until it receives a response from B, it is likely to overload the media and reduce the utilization of the media for other nodes. On the other hand, however, if node A gives up now and tries again later, the probability that node A will succeed in communicating with node B is likely to be severely reduced. This is especially so if node C is very active.

Link Initiation Method

The link initiation by a potential destination node method will be described in more detail. A flow diagram illustrating the link initiation by a potential destination node method of the present invention is shown in Figure 12. In this method, the destination node rather than the source node attempts to capture the media (step 180). Once the media is captured, a link is established with the potential source node (step 182). Thus, the node that considers itself a 'potential blocked destination' of some potential source node initiates the media capture procedure thereby the potential blocked destination plays the roll of the source node and the respective potential source plays the roll of the destination node. In accordance with the invention, a given node (for example node B in the example of Figure 11) considers itself a 'potential blocked destination' if: (1) the media, as seen by the given node (e.g., node B), has been busy during one or more previous frames and (2) either (a) the given node received one or more messages from some other node (e.g., node A), while the media at the given node was busy or (b) the given node determines with high probability that some other node (for example node A in the example of Figure 11) has sent messages to the given node, while the media at the given node was busy but those messages where not received.

Referring to the example of Figure 11, if a repetitive process of sending data messages from node A to node B has been established, than node B may estimate the expected time of arrival of the next message. Examples for such repetitive processes include the delivery of isochronous services such as voice, audio and video data.

When a given node (e.g., node B in the example of Figure 11) considers itself a potential blocked destination, it attempts to capture the media and establish a link with the potential source (e.g., node A). The process of capturing the media comprises the echo based process described herein above. The difference being that normally the source acts as the initiator (i.e. sending the first message to the destination) and the destination node acts as the responder (i.e. sending a response to the source). In the link initiation method described herein, the potential destination acts as the initiator and the potential source acts as the responder.

Release Indication Method The release indication method of the present invention will now be described in more detail. In this method, a node that detects the media to be free, after the media has been busy for a relatively long time, transmits a special release indicator signal. This release indicator functions to signal a potential source, that has failed in its previous attempts to capture the media, to re-attempt capture since there is a high probability that the release indicator signal was transmitted by the potential source's intended destination node.

In accordance with the invention, the release indicator is any suitable signal such as a pulse, which is transmitted in a fixed location at or near the beginning of an available frame. A diagram illustrating the frame structure incorporating the release indication, collision resolution sequence and echo fields inserted into the frame in accordance with the present invention is shown in Figure 13.

Each frame 192 in the stream of frames, generally referenced 190, comprises a header field and a frame body field 198. The header field comprises a busy signaling field

194 and a synchronization field 196. The structure of the body field depends on the media status. If the media is busy, the frame body field 198 is used in an unrestricted way for communication between the nodes occupying the media.

If the media is free, the fame body field 198 comprises, inter alia, a release indication field 200 at a fixed location with the frame body and a collision resolution sequence and echo field 202.

The following is an example of the operation of the release indicator mechanism of the present invention. With reference to Figure 11, node A transmits a message to node B while the media at node B is busy. Consequently, node B does not detect this message because of node C's transmission. After node A fails to communicate with node B, node A backs off for a relatively long time. During this time, however, node A is also searching for a release indication signal.

When nodes C and D eventually release the media, node B sends a release indication at the very first free frame. If the media at node A's location is also free at that frame, node A detects the indication sent by node B and competes for the media at that very frame, having a good probability to succeed (since the media at node B is also free).

A flow diagram illustrating the release indication method of the present invention is shown in Figure 14. To aid in illustrating the principles of the present invention, the method is described in the context of two nodes A and B. Initially, node A transmits a message to node B (step 210). Node B, however, fails to detect the message due to interference or some other reason (step 212). Note A continues until a failure is declared by node A (step 214). At this time, node A backs off for a relatively long time (step 216).

At the same time, node A searches for a release indication (step 218). Once a release indication is detected (step 220), node A attempts to compete for the media (step 222). Once the media is captured, node A establishes a link with node B (step 224). Multiple Priority Media Occupation Method

The multiple priority media occupation method of the present invention will now be described in more detail. In the methods previously described to this point, the occupation of the media occurred within a single state, i.e. the media was either available or busy. When the media is available, any one of a number of nodes may compete for it. When the media is busy, no node can compete for the media until it becomes available again.

The present invention builds on this single level framework by categorizing communication between nodes into several priority levels. When a group of nodes captures and occupies the media, the priority level corresponding to the capture is specified in a specific busy signaling location in the frame. Note that the same nodes can capture the media at different priorities. For example, depending on the type of messages they want to transmit. In addition, the time duration for capture of the media at high priority maybe limited, so the nodes might capture at low priority to enable longer media occupation times. Other nodes that wish to communicate at the same priority level must wait until the media becomes free. The invention, however, provides a mechanism whereby nodes that wish to communicate at a higher priority level may interrupt the current activity and capture the media.

The method incorporates a multi-level busy signaling mechanism and an echo- based interrupt mechanism. For illustration purposes only, the multi-level busy signaling method is presented below within the context of example using two priority levels. It is appreciated that one skilled in the art can extend the method for any number of desired priority levels.

A diagram illustrating the high priority busy, interrupt, interrupt echo and low priority busy fields inserted into the frame in accordance with the present invention is shown in Figure 15. Each frame 232 of the frame stream, generally referenced 230, comprises a header and a body portion 244. The header comprises a plurality of bins (or fields) for performing the multi-level busy signaling mechanism of the invention. In particular, the header comprises four bins including a high priority busy slot 234, an interrupt slot 236, interrupt echo slot 238 and low priority busy slot 240. In addition, the header comprises a synchronization field 242.

The bi-level media occupation mechanism is implemented using the four signals described above. A group of nodes occupies the media at one of the two priority levels. The priority level is indicated in the busy signals transmitted by the nodes, i.e. either a low priority busy signal or a high priority busy signal. When the media is occupied at high priority, the occupation cannot be interrupted. However, when the media is occupied at low priority, it is possible to interrupt the occupation, as described below.

A flow diagram illustrating the multiple priority communication interruption method of the present invention is shown in Figure 16. A node that wishes to compete for the media at high priority transmits a signal at the interrupt bin (step 250). All potential destinations for high-priority capture monitor the interrupt bin for the presence of signal (step 252). In the event an interrupt signal is detected (step 254), the node transmits a pulse in the interrupt echo bin (step 256). Otherwise, the method continues with step 258. Note that a potential destination is a node has the capability of responding to a high priority media capture initiative. In other words, a potential destination is a node that does not detect signal in the high priority busy signal bin.

Next, every node that occupies the media at low priority monitors the interrupt and the interrupt echo bins for the presence of signal (step 258). In the event signal is detected in either the interrupt bin or the interrupt echo bin (step 260), the node responds by releasing the media (step 262). The media can be released by ceasing to transmit low priority busy signals. In the event no signal is detected in either the interrupt bin or the interrupt echo bin (step 260), the node continues operating as normal (step 264).

Note that if a frame has become free due to an interrupt signal, it can be used for media competition at high priority, but not for competing at low priority. A frame is determined to have been freed due to an interrupt signal if (1) signal is not detected at the high priority busy bin and the low priority busy bin, and (2) a signal is detected either at the interrupt bin or the interrupt echo bin.

Note also that if a frame has become free due to other than an interrupt signal, the frame can be used for media competition at any priority level. A frame is determined to have been freed due to other than an interrupt signal if (1) no signal is detected at the high priority busy bin and at the low priority busy bin, and (2) no signal is detected at the interrupt bin and the interrupt echo bin.

Power Level Regulation Method The power level regulation method of the present invention will now be described in more detail. In the invention described in the previous application cited hereinabove, did not include a method for regulating the power level at which the busy signals are transmitted and the threshold at which they are detected. It was implicitly assumed that the busy signals are transmitted at a fixed power level and are detected at a fixed threshold level.

The present invention provides a method to refine and regulate the transmission power level and the detection threshold of the busy signals and to adjust them to the power level of the potential interfering transmitter and to the acceptable interference level at the potentially interfered receiver.

Note that when setting the transmission power level and the detection threshold of the busy signals, the following considerations should preferably be taken into account. Preferably, the busy signal transmitted by a given node should be detected with high likelihood by any node that may potential interfere with the given node. On the other hand, it is desirable that the busy signals not be detected by nodes that are too far to be potential interferers, thus increasing the chance for reuse of the media.

A diagram of a shared network illustrating the range of the busy signal transmitted by a node is shown in Figure 17. In this example network, generally referenced 270, nodes

A and B communicate with each other. It is required that the busy signal transmitted by node B be detected by node C, so as to prevent node C from transmitting and potentially interfering with reception by node B. On the other hand, it is desirable that the busy signal transmitted by node B not be detected by node D because node D is not a potential interferer of node B. This enables nodes D and E to communicate with each other in parallel to the communication taking place between nodes A and B, thus increasing the utilization of the media (this is usually referred to as media reuse).

A flow diagram illustrating the power level regulation method of the present invention is shown in Figure 18. The following describes the method of determining the transmission level detection threshold for the busy signals. In this derivation, it is assumed that node X has captured the media and is transmitting busy signals (step 280) and node Y wants to capture the media and must decide whether the media is busy (step 282). Let PT_XX represent the level at which node X transmits busy signals. Let Attxy represent the path attenuation from node X to node Y. Let P_RXY represent the level at which the busy signals are received by node Y. Note that all values are expressed in dB. Then the level at which the busy signals are received by node Y can be expressed as

"R_XΪ ⁼ "T X ~~ Att_XY (l Further, let P_fXγ represent the level at which node Y intends to transmit. Let Attyx represent the media attenuation from node Y to node X. Let P^x represent the level at which the transmission of node Y will be received at node X. Then PR_XX can be expressed as

Note that the decision by node Y whether the media should be considered busy depends on the value of Atlγχ, which is unknown to node Y. Therefore, node Y estimates Attyx, from the value of Attxy, which node Y can calculate from the value of P^_Y, as follows

Let D represent be maximum expected difference between the attenuation over any given path and the attenuation over a corresponding reverse path. Then

Att_y_ ≤ Att_xγ (3) combining Equations 1, 2 and 3 yields

"RXX — "TXY ~ * TxX ^"1^" "RXY ⁺ " W

Let IMA_X represent the maximum level at which an unwanted received signal can be tolerated and not considered interference. Taking into consideration that transmissions from node Y should not cause interference at the receiver of node X, node Y should be allowed to transmit only if it can be ensured that

"RXX — * MAX )

Assuming that IMA_X, T_XY, PT_XX and D are fixed values, let us define a threshold K as follows (step 288)

K ≡ I_MAX -P_TxY +P_TxX -D (6)

If the power level P_RXY at which the busy signal is detected by node Y does not exceed the threshold K (step 290), it is likely that node Y can transmit at a power level P_Txγ without causing interference to those nodes transmitting busy signals (step 292). If the power level P_RXY at which the busy signal is detected by node Y does exceed the threshold K then node Y declares detection of a busy signal (step 294).

Note that the above derivation refers to the case where IMAX and Pτ_xγ have fixed power levels. In this framework, the busy signals are transmitted at a fixed power level P_TxX and the media is considered busy only if the busy signal is detected at a level PR_XY that exceeds a fixed threshold K. Two improvements to the fixed framework are presented below.

First, the transmission power level P _xγ is a global maximum level. In many cases a lower transmission power level is sufficient for a given node to communicate with its partner node. Such as if the path attenuation is low or if the required transmission rate is low. In these cases, the node sets the threshold for detecting the busy signals at a corresponding lower level. For example, if node Y wishes to communicate at _XY - P_TXY ~ z Λ may set the busy signal threshold at K' = K-z -

Second, the tolerated interference level IMAX is a global minimum level. In many cases the potential interfered node X can tolerate a higher interference level. For example when the noise level at node X is high or when the path loss from its partner node is low. In these cases, node X transmits its busy signals at a corresponding lower power level. For example, if node X is able to tolerate interferences at a level up to T_MAX = I_MAX + z , node X this fact by transmitting busy signals at a power level Bj^ = P_TxX + z ■

Node Device Incorporating the Media Access Controller

An example embodiment of a node device incorporating the media access controller of the present invention will now be described. A block diagram illustrating a node device incorporating a media access controller constructed in accordance with the present invention is shown in Figure 19. The node device, generally referenced 300, represents a node device that may operative stand alone or may be incorporated within a network device such as a switch or router for performing communication functions (i.e. implementing OSI stack protocol functions including MAC functionality). The node device comprises an application processor 308 with associated static memory and dynamic memory (not shown) all in communication with the processor. The application processor is also in communication, via a host interface 316, with a host device 310. The host may be adapted to communicate over one or more other networks.

The node device also comprises media coupling circuitry 302 that functions to physically and electrically interface the node device to the shared media 314. Transmit circuitry 304 and receive circuitry 312 communicate over the media via the media coupling circuitry. The media access controller (MAC) 306 functions, on one side, to provide transmit data to the transmit circuit and to receive data from the receive circuit. On the other side, it interfaces to the application processor. The MAC 306 functions to perform layer 2 (i.e. link layer) functions including controlling access to the media using any combination of the media access control methods described hereinabove, including the echo based collision detection and resolution methods, link initiation methods, release indication methods, multiple level media capture methods and power regulation methods. Note that the media access control methods described may be implemented in software and adapted to reside on a computer readable medium, such as a magnetic disk, floppy disk, Flash memory card, EEROM based memory, bubble memory storage, RAM storage, ROM storage, etc. The software may also reside, in whole or in part, in the static or dynamic main memories or in firmware within the processor of a computer system (i.e. within microcontroller, microcomputer internal memory). The host or other network interface device is capable of transmitting and receiving a carrier wave signal which encodes the media access control software. In particular, the media access control software comprises a sequence of instructions which, when executed by the processor, cause the computer system to perform the steps of any of the methods described hereinabove. In alternative embodiments, the present invention may be applicable to implementations of the methods and apparatus described above in integrated circuits, especially Application Specific Integrated Circuits (ASICs) or chip sets, Field Programmable Gate Arrays (FPGAs), wireless modem implementations, powerline modem implementations, switching system products and transmission system products. Note that a combination of software and hardware can also be implemented, the former performing the complex operations and the latter performing the time critical operations.

It is intended that the appended claims cover all such features and advantages of the invention that fall within the spirit and scope of the present invention. As numerous modifications and changes will readily occur to those skilled in the art, it is intended that the invention not be limited to the limited number of embodiments described herein. Accordingly, it will be appreciated that all suitable variations, modifications and equivalents may be resorted to, falling within the spirit and scope of the present invention.

Claims

1. A method for detecting collisions among a plurality of competing source nodes in a distributed network of nodes connected to a media, said method comprising the steps of: each competing node transmitting a collision detection sequence onto said media; each non-competing node monitoring said media and listening for a collision detection sequence; each non-competing node echoing the collision detection sequence heard thereby; each competing node listening for said echoed collision detection sequence and comparing its originally transmitted collision detection sequence therewith; backing off from transmission if the echoed collision detection sequence does not match the collision detection sequence previously transmitted; and transmitting a message onto said media if the echoed collision detection sequence does match the collision detection sequence previously transmitted.

2. The method according to claim 1, wherein said collision detection sequence is transmitted by said competing nodes in a collision detection sequence field at a specific location within each frame.

3. The method according to claim 1, wherein said echoed collision detection sequence is transmitted by said non-competing nodes in an echo collision detection sequence field at a specific location within each frame.

4. The method according to claim 1, wherein said echo collision detection sequence field is located at a position within said frame later in time than said collision detection sequence field.

5. The method according to claim 1, wherein the collision detection sequence transmitted by each competing node is unique among said competing nodes.

6. The method according to claim 1, wherein the collision detection sequence transmitted by each competing node is selected randomly by each competing node.

7. The method according to claim 1, further comprising the step of each competing node listening to said media during the off bins of its collision detection sequence, and backing off from transmission if signal is detected during an off bin.

8. The method according to claim 1, wherein an echoed collision detection sequence is declared not to match the collision detection sequence previously transmitted if signal is detected in a bin within said echoed collision detection sequence corresponding to an off bin within said collision detection sequence.

9. The method according to claim 1, further comprising the steps of: one or more destination nodes competing for said media wherein each competing destination node transmitting a collision detection sequence and listening during off bins; and if no collision is detected by a competing destination node, responding to the source node and proceeding with communications between said source node and said destination node; and if a collision is detected by a competing destination node, ceasing transmitting and not responding to the source node.

10. The method according to claim 9, wherein the collision detection sequence transmitted by the competing destination node differs from the collision detection sequence transmitted by the source node.

11. A method for resolving collisions among a plurality of competing source nodes in a distributed network of nodes connected to a media, said method comprising the steps of: each competing node transmitting a collision detection sequence one bin at a time onto said media; each non-competing node monitoring said media and listening for a collision detection sequence bin; each non-competing node echoing the collision detection sequence bin heard thereby interleaved with the transmission of said collision detection sequence; each competing node listening for said echoed collision detection sequence bin and comparing the collision detection sequence bin previously transmitted thereby; backing off from transmission if the echoed collision detection sequence bin does not match the collision detection sequence bin previously transmitted; and transmitting a message onto said media if the echoed collision detection sequence bin does match the collision detection sequence bin previously transmitted.

12. The method according to claim 11, wherein the competing node having the largest collision detection sequence, when expressed as a binary number, succeeds in capturing the media.

13. The method according to claim 11, wherein said collision detection sequence bins are transmitted by said competing nodes at specific locations within each frame.

14. The method according to claim 11, wherein said echoed collision detection sequence bins are transmitted by said non-competing nodes at specific locations within each frame.

15. The method according to claim 11, wherein the collision detection sequence transmitted by each competing node is unique among said competing nodes.

16. The method according to claim 11, wherein the collision detection sequence transmitted by each competing node is selected randomly by each competing node.

17. The method according to claim 11, further comprising the step of each competing node listening to said media during the off bins of its collision detection sequence, and backing off from transmission if signal is detected during an off bin.

18. The method according to claim 11, wherein an echoed collision detection sequence bin is declared not to match the collision detection sequence bin previously transmitted if signal is detected in the echoed collision detection sequence bin corresponding to a off collision detection sequence bin.

19. The method according to claim 11, further comprising the steps of: one or more destination nodes competing for said media wherein each competing destination node transmitting a collision detection sequence and listening during off bins; and if no collision is detected by a competing destination node, responding to the source node and proceeding with communications between said source node and said destination node; and if a collision is detected by a competing destination node, ceasing transmitting and not responding to the source node.

20. The method according to claim 19, wherein the collision detection sequence transmitted by the competing destination node differs from the collision detection sequence transmitted by the source node.

21. A method for initiating a link between a potential source node and a blocked destination node in a distributed network of nodes connected to a media, said method comprising the steps of: determining that a node is a potential blocked destination node; the blocked destination node attempting to capture the media; and once the media is captured, establishing a connection between said destination node and said source node.

22. The method according to claim 21, wherein a node is determined to be a potential blocked destination node when it sees the media as busy during at least the previous frame and the node detects that it has received one or more messages from another node.

23. The method according to claim 21, wherein a node is determined to be a potential blocked destination node when it sees the media as busy during at least the previous frame and the node detects that it is likely that another node transmitted messages to it during the time said media was busy.

24. The method according to claim 21, wherein said step of capturing a node comprises the steps of: each non-competing node monitoring said media and listening for a collision detection sequence; each non-competing node echoing the collision detection sequence heard thereby; each competing node listening for said echoed collision detection sequence and comparing its originally transmitted collision detection sequence therewith; backing off from transmission if the echoed collision detection sequence does not match the collision detection sequence previously transmitted; and transmitting a message onto said media if the echoed collision detection sequence does match the collision detection sequence previously transmitted.

25. A method for indicating the release of a media by a node in a distributed network of nodes connected to said media, said method comprising the steps of: a node detecting that the period of time said media has been busy exceeds a threshold; detecting that said media has become available; transmitting a release indication signal onto said media indicating that the media is now available; and a potential source node detecting said release indication signal and in response thereto attempting to capture said media.

26. The method according to claim 25, wherein said threshold is predetermined.

27. The method according to claim 25, wherein said threshold is set dynamically by each node.

28. The method according to claim 25, wherein said potential source node has repeatedly failed in previous attempts to capture said media.

29. The method according to claim 25, wherein said release indication signal comprises a pulse transmitted at specific locations with each frame.

30. The method according to claim 25, wherein said step of detecting that said media has become available comprises checking for the presence of a signal in a frame occupation field within each frame.

31. A method of prioritized media capture for use in a distributed network of nodes connected to a media, said method comprising the steps of: transmitting a low priority busy signal when a node occupies the media at low priority; transmitting a high priority busy signal when a node occupies the media at high priority; transmitting an interrupt signal when a node wants to compete for the media at high priority; each potential high priority destination monitoring the media for the presence of an interrupt signal; and if detected, echoing said interrupt signal as an interrupt echo signal; each node occupying the' media at low priority monitoring the media for the presence of either said interrupt signal or said interrupt echo signal; and in response to the detection thereof, releasing the media.

32. The method according to claim 31, wherein the step of releasing the media comprises ceasing to transmit low priority busy signals.

33. The method according to claim 31, further comprising the step of using a frame that has become available due to an interrupt for media competition at high priority.

34. The method according to claim 33, wherein it is determined that a frame becomes available due to an interrupt if no low priority busy signal or high priority busy signal is detected and either an interrupt signal or an interrupt echo signal is detected.

35. The method according to claim 31, further comprising the step of using a frame that has become available by other than an interrupt for media competition at any priority.

36. The method according to claim 35, wherein it is determined that a frame becomes available due by other than an interrupt if neither a low priority busy signal, high priority busy signal, interrupt signal nor an interrupt echo signal is detected.

37. The method according to claim 31, wherein said low priority busy signal is transmitted at a specific location within each frame.

38. The method according to claim 31, wherein said high priority busy signal is transmitted at a specific location within each frame.

39. The method according to claim 31, wherein said interrupt signal is transmitted at a specific location within each frame.

40. The method according to claim 31, wherein said interrupt echo signal is transmitted at a specific location within each frame.

41. A method for dynamically regulating the power level of busy signals transmitted by nodes in a distributed network of nodes connected to a media, said method comprising the steps of: adjusting the transmission power of the busy signals transmitted by a node to a level such that the busy signals are detected by nodes potentially interfering with said node while they are not detected by nodes sufficiently distant from said node to be non-potentially interfering nodes; and adjusting the detection threshold of busy signals received by said node such that the busy signals transmitted by potentially interfering nodes are detected by said node while the busy signals transmitted by non-potentially interfering nodes are not detected by said node.

42. The method according to claim 41, wherein a node detects a busy signal if the power level of said busy signal exceeds a threshold.

43. The method according to claim 41, wherein a node detects a busy signal if the power level of said busy signal exceeds a threshold K given by

& ~ '^■ MAX ~~ "T Y ^■*^" "TXX ~ *-*

where IMA_X is the maximum level at which an unwanted signal tolerated and not considered interference;

P_TXY is the level at which a competing node Y intends to transmit; PT_XX is the level at which a node X that has captured the media transmits busy signals; D is the maximum expected different between the attenuation over any given path and the attenuation over the reverse path.

44. The method according to claim 41, further comprising the step of a node indicating that is able to tolerate interference at a level higher than IMAX by correspondingly reducing the transmission power level of the busy signals.

45. The method according to claim 41, further comprising the step of a node reducing its threshold for detection of the busy signals when it intends to transmit at a power level lower than Pτ_xγ .