GB2455792A - Allocating resources in networks using directional antennas - Google Patents

Abstract

A method and apparatus for contending for medium access in networks using directional antennas. Apparatus in a network are operable to transmit and receive control packets in all available directions within a synchronised access period, so that neighbouring nodes are made aware of an intended directional data communication during a subsequent data transmission period. From one or more of these control packets, an apparatus can estimate the likelihood of interference between a directional communication that it intends to participate in and other intended and ongoing directional communications in the network.

Description

1 2455792 The invention is concerned with a method and apparatus for allocating resources in networks using directional antennas. The invention can be employed particularly, but not exclusively in 60 GHz wireless personal area networks.

High data rate communications in the 60 GHz band have received considerable interest, especially for use in the home network area. Ongoing standards development efforts in this area include proposals by the IEEE 802.15.3c Millimeter Wave Alternative PHY Task Group (TG3c) (http://www.ieee802.org/l5/pub/TG3c.html) and the WirelessHDTM Special Interest Group (http://www.wirelesshd.org).

In addition to the potential of extremely high data rate (several Gbps) to support applications such as uncompressed High-definition television (HDTV) and video-on-demand, one advantage of 60 GHz systems is that directional antennas are much easier to implement than at 5 GHz. This is because of the smaller wavelength and smaller fractional bandwidth (see, for example, N. Guo et al., "60 GHz millimetre-wave radio: principle, technology, and new results", Journal on Wireless Corn. and Networking, 2007).

Past work has shown that, in terms of signal-to-noise ratio (SNR) at a receiver, directional antennas bring two major benefits, namely increased spatial reuse and improved signal quality. A study by Li et al (G. Li et al., "Opportunities and challenges for mesh networks using directional antennas", IEEE WiMesh, 2005) suggests that, as a result, directional antenna technology can lead to a capacity/throughput improvement of two or four times depending on the topology. Moreover, directional antennas have higher gains than omm-directional antennas and hence have a longer transmission range. Therefore, networks using directional antennas are expected to have improved routing performance and better network connectivity.

On the other hand, the 60 GHz band currently requires line of sight (LOS) between transmitter and receiver. Furthermore, the capability of 60 GHz transmissions to diffract around obstacles is much less than that of 5 0Hz, for example, thus limiting its range.

Moreover, higher frequency results in higher path loss.

It is envisaged that nodes operating in accordance with high data rate communications in the 60 0Hz band can be equipped with a directional antenna consisting of N antenna elements, where each antenna element has an azimuthal radiation pattern spanning an angle of 27tfN radians. An exemplary antenna system of six elements, which collectively cover the entire plane (360 degrees), is shown in Figure 1.

Such nodes could further be expected to exhibit at least some of the following features: the ability to maintain the orientation of the antenna elements at all times, irrespective of node movement; the use of a Medium Access Control (MAC) protocol able to select electronically any one or all of the antenna elements for transmitting a signal; when receiving a signal, a node can use the signal from the antenna element that has the maximum power of the desired signal; and, the use of only one transceiver (radio interface) at each node. The latter feature is considered to constitute a bottom-line' scenario, since multiple radios enable much simpler protocols, in part because problems such as deathess and hidden nodes can be solved more easily.

As the MAC protocol for 60 GFIz networks has not yet been standardized, networks described hereinafter are discussed with reference to existing MAC protocols, such as 802.11, 802.15.3, or WiMedia, though the present invention is not limited thereto.

Furthermore, it will be appreciated that network codes can be fully distributed (as in 802.11 ad hoc mode or WiMedia), or there can be a central node (e.g. an Access Point, AP).

In conventional 802.11 or 802.15.3 networks, omni-directional transmission and reception of beacons or HELLO messages are used to identify neighbours. In a network with directional antennas, these messages are still transmitted (e.g., by an AP) 0mm-directionally (i.e. on all antenna elements). Upon receiving such a message, a station records the direction of the sender by noting the antenna element that received the maximum power. It then uses the antenna element towards that specific direction to send back a response message. After the discovery process, each node has built up a topology lookup table containing two fields: NodelD and Antenna, the latter being the antenna element used to beamform towards the node with the NodeID.

There are generally two types of MAC: contention-based and contention free. The first type of protocol includes IEEE 802.11 DCF and 802.15.3 contention access and is currently the dominant MAC in use. The second type of protocol includes 802.1 le HCCA and 802.15.3 channel time allocation, where channel access is centrally allocated in a TDMA fashion.

For contention-based MAC protocols in networks implementing directional communications, proposals in the art are similar to basic IEEE 802.11 DCF or 802.lle EDCA, with adaptations for directional antennas (R. Choudhury et a!., "Using directional antennas for medium access control in ad hoc networks", ACM Mobicom, 2002). Channel reservation is performed using a Request-to-Send/Clear-to-Send (RTS/CTS) handshake between a sender S and a receiver D. When S wants to send a data packet to D for the first time there are two Qptions: ifS already knows the direction of D (from the lookup table), it transmits a RTS directionally to D; otherwise, it transmits a RTS to 1) on all antenna elements. Node D receives the RTS intended for it and estimates the direction from which it received the RTS. It checks its lookup table to see if there is an entry of S. If not, it records NodelD and Antenna of S. On the other hand, if there is already an entry of S but with a different direction (antenna), it updates the entry with the new direction since relative positions between nodes may change due to mobility. D then uses the same antenna element with which it received the RTS to send back a CTS. Similarly, S estimates the direction of D while receiving the CTS packet, and updates this information in its own lookup table. After the RTS-CTS handshake is performed successfully, S starts transmitting data packets using the antenna element pointing towards D. In the case of 802.1 le EDCA (enhanced distributed channel access), S can send multiple frames within a TXOP (transmission opportunity).

Nodes in the neighbourhood of S and D which overhear the RTS-CTS dialogue defer transmission for the proposed duration of transfer using a NAV (network allocation vector) table, similar to 802.11. An enhancement for directional antennas is that these nodes defer transmission only for the antenna element with which RTS or CTS is overheard (R. Choudhury et al., "Using directional antennas for medium access control in ad hoc networks", ACM Mobicom, 2002). In this case, each node maintains a directional NAY table comprising the fields Antenna and Blocked. On overhearing a RTS or CTS on antenna element i, a node sets the Blocked flag for element i, for the proposed duration. Subsequently, this node defers transmissions which requires use of element i. After the proposed duration is over, the Blocked flag is reset.

One of the problems with the use of directional antennas in such MAC protocols is deafness. A node using a directional antenna is considered deaf in all directions except that of its main beam. Figure 2 illustrates an example of a scenario that results in deafness. Node A is engaged in communication with Node B using beamforming in the direction of B. Node C, unaware of this on-going transmission, attempts to transmit data to A. Since the main lobe of A's antenna is not in the direction of Node C, it does not hear' RTS packets sent by C, which results in timeouts in Node C. The consequences of the deafness problem can further propagate throughout the network and cause low performance, unfairness, and deadlock.

A related problem is the presence of hidden nodes caused by unheard RTS/CTS.

Consider the example scenario depicted in Figure 3. Beamforming enables the existence of two simultaneous transmissions: from A to B and from C to D. Assume that when C wants to communicate with D, nodes A and B are already communicating, with A's antenna beam 302 pointing towards B. Even if C and D exchange RTS/CTS omni-directionally, A will not hear it. Now suppose that the communication of A and B ends first and node A has a packet for node D. In this scenario, if A tries to send a RTS packet to node D, the RTS packet may collide with the data packet sent by C (indicated by the overlap of transmission patterns 304 and 306). This kind of hidden node problem arises because a node misses an RTS/CTS exchange in its neighbourhood and initiates a transmission to the receiver of an ongoing transmission.

Aspects of the present invention seek to mitigate or eliminate at least some of the above-mentioned problems or disadvantages.

In general terms, an aspect of the present invention provides that nodes in the network are operable to exchange control packets, by which they register their intention of participating in data communication, during a synchronised time interval. An exemplary time interval of this kind is the Ad hoc Traffic Indication Message (ATIM) window in the IEEE 802.11 power saving mechanism (PSM), discussed in more detail below.

Combined with a capability to exchange control packets omni-directionally, better organization of the future use of a channel for directional data transmission is achieved.

In IEEE 802.11 PSM, a node can save energy by going into a dormant state (doze mode). In doze mode, a node consumes much less energy compared to normal mode, but cannot send or receive packets. It is desirable for a node to enter a dormant state only when there is no need for exchanging data. In IEEE 802.11 PSM, this power management is done based on Ad hoc Traffic Indication Messages (ATIMs). Time is divided into beacon intervals, and every node in the network is synchronized by periodic beacon transmissions. In this way, every node will start and finish each beacon interval at about the same time.

Figure 4 illustrates the process of IEEE 802.11 PSM. At the start of each beacon interval, there exists a time interval called the AIIM window where every node should be in a wake state. If node A has buffered packets for B, it sends an ATIM packet to B during this time interval. If B receives this message, it replies back by sending an ATIM-ACK to A. Both A and B will then stay awake for that entire beacon interval. If a node has not sent or received any ATIM packets during the ATIM window (e.g., node C), it enters doze mode until the next beacon time.

In the academic literature (J. So and N. Vaidya "Multi-Channel MAC for Ad Hoc Networks: Handing Multi-Channel Hidden terminals Using A Single Transceiver", ACM MobiHoc, 2004), ATIM windows are used to solve the frequency formed multi-channel hidden node problem. By contrast, the present invention makes use of a periodically recurring window, such as the ATIM window, as a solution to the deafness and directional hidden node problems.

Therefore, according to a first aspect of the invention, there is provided a transceiver for use in a wireless communication network, the transceiver comprising: means for defining one or more access periods and respective subsequent data transmission periods such that said periods are synchronised with those of other nodes in the network; means for transmitting and receiving communication packets in any one or more of a plurality of directions, said means being adapted to transmit and receive control packets in all available directions during an access period; and means for contending for network medium access during an access period, said means for contending being operable to determine a likelihood of interference between a directional data packet communication in which the transceiver intends to participate in during a subsequent data transmission period and other intended or ongoing directional data packet communications in the network, wherein said likelihood is determined in consideration of any control packets received by the transceiver.

According to a second aspect of the present invention, there is provided a method of contending for medium access at a transceiver for use in a wireless communication network, the method comprising: defining one or more access periods and respective subsequent data transmission periods such That said periods are synchronised with those of other nodes in the network, the transceiver being adapted to transmit and receive control packets in all available directions during said access period; and determining, in consideration of any control packets received by the transceiver, a likelihood of interference between a directional data packet communication in which the transceiver intends to participate in during a data transmission period and other intended or ongoing directional data packet communications in the network.

According to a third aspect of the invention, there is provided a wireless communication network comprising a plurality of transceivers each according to the first aspect of the invention.

According to a fourth aspect of the invention, there is provided a method of communicating in a wireless communication network having a plurality of transceivers, the method comprising: contending for medium access at each transceiver intending to participate in a directional data communication according to the second aspect of the invention; and implementing said intended participation at each transceiver determining that the likelihood of interference between a directional data packet communication in which the transceiver intends to participate in and other intended or ongoing directional data packet communications in the network is low, and in accordance with a successful control packet exchange.

According to a fifth aspect of the invention, there is provided a computer program for implementing an apparatus according to the first aspect of the invention.

According to a sixth aspect of the invention, there is provided a carrier medium carrying processor executable code for controlling a processor to carry out the method of the second aspect of the invention.

These and other aspects of the invention will now be further described, by way of example only, with reference to the accompanying figures, in which: Figure 1 depicts the azimuthal radiation pattern of node having six antenna elements; Figure 2 schematically illustrates the deafness problem in a network of nodes; Figure 3 schematically illustrates the hidden node problem in a network of nodes; Figure 4 schematically illustrates the use of ATIM windows at the beginning of a beacon period in the MAC structure of IEEE 802.11 PSM; Figure 5 depicts the azimuthal radiation pattern of a node according to the present invention; Figure 6 shows a schematic of a node structure according to the present invention; Figure 7 schematically illustrates the use of time windows in directional MAC according to the present invention; Figure 8 schematically shows a first network comprising a plurality of nodes in accordance with the present invention; Figure 9 schematically shows a second network comprising a plurality of nodes in accordance with the present invention; Figure 10 is a flow diagram showing the operation of a data transmitting node contending for channel access in accordance with the present invention; Figure 11 is a flow diagram showing the operation of a data receiving node contending for channel access in accordance with the present invention; Figure 12 schematically shows a third network comprising a plurality of nodes in accordance with the present invention.

An exemplary wireless network according to the present invention comprises a plurality nodes, which is a generic term referring to a point of communication in a network. Each node may comprise a transceiver configured in accordance with the depictions of figures 5 and 6.

Figure 5 depicts the azimuthal radiation pattern of an exemplary node 500 equipped with a multi-directional antenna comprising four antenna elements. Each antenna element of the node covers one of four 90 degree sectors in the azimuth direction, such that they collectively cover the entire plane (360 degrees). For convenience, antenna elements are referred to in the context of the sector in which the main lobe of their corresponding transmission pattern 501, 502, 503, 504 extends. It will be appreciated that a node has the option to use all antenna elements simultaneously, resulting in omni-directional transmission/reception.

In figure 6, an embodiment of transceiver 600 according to the present invention comprises a processor 602 operable to execute machine code instructions stored in a working memory 604 and/or retrievable from a mass storage device 606. By means of a general-purpose bus 608, user operable input devices 610 are in communication with the processor 602. The user operable input devices 610 comprise any means by which an input action can be interpreted and converted into data signals, for example, DIP switches.

Audio/video output devices 612 are further connected to the general-purpose bus 608, for the output of information to a user. Audio/video output devices 612 include any device capable of presenting information to a user, for example, status LEDs.

A communications unit 614 is connected to the general-purpose bus 608, and further connected to an antenna or set of antennas 616. By means of the communications unit 614 and said antenna 616, the transceiver 600 is capable of establishing wireless communication with other nodes. The communications unit 614 is operable to convert data passed thereto on the bus 608 to an RF signal carrier in accordance with a communications protocol previously established for use by a system in which the transceiver 600 is appropriate for use, for example WirelessHD.

In the transceiver 600 of figure 6, the working memory 604 stores applications 618 which, when executed by the processor 602, cause the establishment of an interface to enable communication of data to and from other nodes. The applications 618 thus establish general purpose or specific computer implemented utilities and facilities that are used in linking nodes.

Figure 7 illustrates, in a time line format, an access contention mechanism in accordance with the present invention. A beacon interval 702, representing the amount of time between beacon 708 transmissions from nodes A, B, C, D in a network, is divided into a first time interval 704 and a second time interval 706. The first time interval 704, hereinafter also referred to as an access window or access period, may comprise an ATIM window or the like.

In this network, node A intends to transmit a data packet 710 to node B, while node C intends to transmit a data packet 711 to node D. Therefore, during the first time interval 704, each of nodes A, C initiates a two-way handshake with its intended destination node B, D. The handshake comprises the transmission of an RTS packet 712, 713 in all available transmission directions (i.e. from all available antenna elements), thereby implicitly directing all nodes within range to back off from transmitting during the data transmission period. However, since nodes in the network are adapted to selectively transmit data in any one of a plurality of transmission directions, a node overhearing a RTS is operable to defer data transmission only for the antenna element with which it receives the RTS, in order to avoid needlessly silencing all of the antenna elements.

A destination node B, D that receives an RTS intended for it replies with an omni-directional CTS packet 714, 715. The CTS similarly has the potential to silence neighbour nodes, while at the same time giving the source the go-ahead to transmit the data packet. Nodes overhearing either control packet update their DNAV tables accordingly.

Upon successfully completing their RTS/CTS handshakes, nodes A, C can send their data 710, 711 to nodes B, D during the second time interval 706. Nodes that are unsuccessful in making a reservation enter a defer state during this time interval 706.

While in a defer state, a node must at least refrain from (re)transmitting control packets.

Typically, the destination nodes B, D acknowledge receipt of the data 710, 711 by means of an acknowledgement (ACK) packet 716, 717.

Since data communication is performed in packet format, a transmission can be performed such that it does not break transmission window limits, even if its intended duration exceeds the duration of a transmission window. For example, the transmission can be spread across a number of transmission windows punctuated by periods of non-transmission during intervening access windows.

It will be apparent that synchronization is required so that all the nodes begin their beacon interval 702 at the same time. To this end, a GPS or the IEEE 802.11 timing synchronization function (TSF) can be used.

When multiple nodes start sending control packets at the beginning of a beacon interval, control packets may collide with each other. To avoid such collisions, each node can wait for a random back-off interval 718, 719 (e.g. chosen from [0, W], where Wis a back-off counter) before transmitting a RTS packet. Since the transmission of control and data packets is separated in time, nodes do not miss the RTS/CTS exchange in their neighbourhood, thereby preventing the directional hidden node problem.

The choice of access window 704 size is a trade-off. When there are a small number of flows in the network, using a large access window is wasteful. Much of the time the channel will be left as idle, because data packets are not allowed to be transmitted in this interval. On the other hand, if there are very large numbers of flows in the network, a longer access window would be needed to exchange all the control messages between nodes. Thus the access window size would affect the network throughput. Ideally access window sizes should be adapted dynamically to the number of flows and network conditions, such as bandwidth, latency, and jitter.

Exemplary situations where the hidden node and deafness problems are mitigated in networks in accordance with aspects of the present invention will now be discussed with reference to figures 8 and 9. The nodes depicted therein are configured in accordance with the embodiments described above. For clarity, not all of the transmission patterns are shown.

Figure 8 shows a network 800 comprising four nodes 802, 804, 806, 808. Node 802 is in radio coverage with node 804, node 804 with node 802 and node 806, node 806 with node 804 and node 808, and node 808 only with node 806. However, only the transmission patterns corresponding to antenna elements 1, 3 of nodes 804, 806 are depicted. Node 804 and node 806 are contending for network access during the same transmission window: node 804 intends to transmit data to node 802, while node 806 intends to transmit data to node 808. All of the nodes are synchronised in time to be in an omni-directional transmission/reception mode during the corresponding access window.

From the omni-directional RTS/CTS exchange between node pair (804, 802) and node pair (806, 808), each initiator 804, 806 hears the other's RTS 810, 812, and sets its directional NAV table accordingly. If nodes 804, 806 do not have knowledge of the local network topology, they can determine the angular location of their neighbour from RTS angle of arrival estimations. In this way, each node 804, 806 keeps track of the direction, and corresponding duration, toward which it must not initiate a data transmission during one or more transmission windows. In this particular instance, node 806 sets a Blocked flag for element I and node 804 sets a Blocked flag for element 3.

Neither initiator 804, 806 hears the omni-directional CTS originating from the other's responder 808, 802, since they are out of range. In such situations, and if the RTS/CTS handshakes are successfully completed, the initiator nodes 804, 806 can assume that they are exposed nodes and are permitted to transmit their intended data transmissions to their respective nodes 802, 808 simultaneously.

Equally, it will be apparent that, in this particular network topology, if node 804 is an intended data recipient of node 802, node 806 can estimate the likelihood of interference of its intended transmission to node 808 based only on a received CTS from node 804. Meanwhile, if node 808 moves to a position that is within the radio coverage of node 804, as indicated by position 808', node 804 can estimate the likelihood of interference on the basis of both the RTS from node 806 and the CTS from node 808'.

In figure 9, the radio coverage in network 900 is such that nodes 902, 906 are in coverage with every other node, while nodes 904, 908 are in coverage with nodes 902, 906, but not with each other. As in the previous example, only a selection of transmission patterns is shown.

Node 906 intends to transmit data to node 904, while node 908 intends to transmit data to node 902, both during the same transmission window. If the random back-off value, W, is lower for node 906 than for node 908, node 906 commences its RTS/CTS exchange first. Both omni-directional transmissions of the (assumedly successful) exchange are overheard by node 902; that is, the RTS 910 from node 906 on antenna element 3 and the CTS (not shown for clarity) from node 904 on antenna element 4.

Since node 908 does not hear the CTS from node 904, and if node 908 does not know the relative position of node 904, the identification of which is determinable from node 906's RTS, node 908 may still contend for network access. However, if node pair (908, 902) were also to successfully complete its RTS/CTS exchange, it is likely that interference between the two ensuing directional data transmissions would arise, which would be particularly noticeable by node 902.

Accordingly, node 902 suppresses its omni-directional CTS reply to node 908 on the basis that at least one of the control packets (the RTS from node 906) was overheard on the antenna element 3 with which it intends to receive a directional data transmission from node 908. Similarly, and referring again to figure 8, node 808' suppresses its CTS transmission to node 806 since it overhears the RTS 810 from node 804 using the same antenna element 1 with which it intends to receive a directional data transmission from node 806.

Likewise, if node pair (908, 902) performs its control packet exchange first, node 906 will overhear the dialogue on antenna elements 2, 1, estimate a high likelihood of interference, and defer transmission of its RTS to node 904 for the duration specified in node 908's RTS.

It will be apparent that if node 902 moves to a position 902' that is outside the radio coverage of 906 andlor node 904 moves to a position 904' that is outside the radio coverage of 902, both intended data transmissions can occur simultaneously. In this way, the access contention process of the present invention avoids needlessly reducing spatial reuse in a network where the implementation of directional transmissions does not lead to collisions.

In the exemplary situations described above, it is assumed that nodes do not have knowledge of the directions of forthcoming data communications. That is, nodes receiving only one RTS or CTS packet may be unaware from which antenna element the node transmitting the RTS or CTS intends to transmit/receive directional data.

However, this need not always be the case. For example, and referring again to figure 8, node 808' may be excluded from automatically suppressing its CTS transmission to node 806 if, by receiving the control packet from node 804, it can derive the necessary directional information corresponding to node 804's intended directional data communication. Exemplary ways in which such information could be derived are detailed below.

Each node may have knowledge of its local topology, in the form of a stored lookup table. In figure 9, node 908 could estimate the likelihood of interference by cross-referencing the directional information stored in its lookup table with the identifiers/addresses of nodes 906, 904 contained in the RTS 910. As before, if node 908 estimates a high likelihood of interference, it must defer transmission of its RTS to node 902 for the duration specified in node 906's RTS.

Alternatively still, in order to decide whether an intended communication will interfere with other intended or on-going communications, neighbouring nodes may be informed about the direction of a data transmission, or the relative location of a node, from information contained in the control packet. For example, the RTS 910 from node 906 may include a directional indicator indicative of the direction of the intended data transmission, or a positional indicator indicative of the position in the network of node 906.

However, this may require the availability of a common reference direction to be used to measure the transmission angle, or a common coordinate system to be used to estimate a relative position, both of which are not necessarily realistic assumptions. The inclusion of such information may also increase control packet size. Therefore, there is clearly a trade-off.

Methods of transceiver operation in accordance with the present invention will now be discussed with reference to figures 10 and 11. Figure 10 is a flowchart representing operation of a transceiver intending to transmit data, while figure 11 is a flowchart representing operation of a corresponding transceiver intending to receive the transmitted data. It is assumed that neither transceiver is participating in an ongoing data communication or intends to participate in another data communication.

In figure 10, the method commences with the transceiver entering the access period (step S 1002), in synchronicity with other nodes in the network. In order to avoid potential control packet collisions, the transceiver then waits for a random back-off period (step S 1004), during which period it is also operable to listen for 0mm-directional RTS and CTS packets transmitted from other nodes in the network (step S 1006).

Once the back-off period expires, and provided that the likelihood of interference between its intended transmission and other intended or ongoing transmissions in the network is estimated to be small (step S 1008, No'), the transceiver is free to contend for network access. The likelihood is estimated in accordance with the foregoing examples. Subsequently, in step S 1010, the transceiver transmits its own RTS on all available antenna elements (i.e. omni-directionally), to alert neighbouring nodes.

Having transmitted its RTS, the transceiver then waits for the corresponding CTS packet. If the transceiver fails to secure channel access, either because it receives no CTS (step S1012 No') or because it estimates that there is a likelihood of interference between its intended transmission and another transmission (step Si 008, Yes'), it enters a defer state (step S 1016) during the transmission period. While in a defer state, the transceiver must refrain from (re)transmitting its RTS control packet. Otherwise, the transceiver transmits its data directionally to the destination node (step S 1014) and waits for an acknowledgement (step S 1018).

The complimentary operation of a receiving transceiver, depicted in figure 11, similarly commences with an entrance into the access period (step 1102). Since this transceiver does not intend to transmit data, it listens for omni-directional control packets from neighbouring nodes (step S 1104). Assuming that it receives the RTS intended for it from the transceiver of figure 10 during the access period, the transceiver decides whether to reply with a CTS (step S 1106), in accordance with the aforementioned exemplary decision making scenarios. If it is unable to participate in data communication (step Si 106, No'), it transitions to a defer state during the transmission period (step S 1110). While in this state, the transceiver must refrain from attempting to (re)transmit its CTS. Otherwise, upon successflully handshaking, it waits for the data transmission from the source (step S 1108), and transmits a directional acknowledgement in response (step SIl 12).

It will be understood that the flowcharts of figures 10 and 11 present the basic operations of transceivers in accordance with the present invention, and that further processes known to those skilled in the art can be incorporated. These include, but are not limited to, multiple transmission attempts during the access window, and dropping a pending intended transmission or reception in favour of another intended transmission or reception.

The discussion now briefly focuses on contention-free MAC protocols in networks making use of directional antennas, such as 802.1 le Hybrid Coordination Function (HCF) Controlled Channel Access (HCCA) and 802.15.3 Channel Time Allocation (CTA).

In these protocols, communications are scheduled instead of contended, so deafness does not occur (see, for example, IEEE 802.1 le WG, "Wireless medium access control (MAC) and physical layer (PHY) specifications: medium access control (MAC) enhancements for quality of service (Q0S)"; and IEEE P802.llefDraft 6.0, November 2004, and IEEE 802.15.3 Working Group, "Part 15.3: wireless medium access control (MAC) and physical layer (PHY) specifications for high rate wireless personal area networks (WPAN)", Draft Standard, June 2004).

Using 802.lle HCCA as a starting point, figure 12 depicts an exemplary network 1200 including an access point 1202, such as a Quality of Service (Q0S)-enhanced access point (QAP), and a plurality of stations 1204, 1206, 1208, 1210 comprising Quality of Service (QoS)-enhanced stations (QSTAs).

In an enhancement of the known protocol, stations 1204, 1206, 1208, 1210 learn the direction of the access point 1202 from the beacons that it sends out and choose an appropriate antenna element for communication. From the associated responses sent by each of the stations, the access point also learns the direction of the stations in its cell and records this information in a lookup table. In this way it builds up a topology map of the cell. This needs updating regularly since stations may be moving, such the movement of station 1208 to position 1208'.

In order to initiate a traffic stream (TS) connection, a station, say station 1204, sends a QoS request frame containing a traffic specification (TSPEC) to the access point 1202.

A TSPEC describes the Q0S requirements of a IS, such as mean/peak data rate, mean/maximum frame size, delay bound, and maximum required service interval (RSI).

From the TSPEC it received, the access point 1202 can estimate the direction of station 1204 and update its lookup table accordingly if needed. In effect, the access point partitions the network into different sectors 1212, 1214, 1216 according to different antenna elements (i.e. different directions), with each sector covering anumber of stations. Using directional antennas can effectively reduce interference among different sectors.

On receiving all the Q0S requests, the access point node 1202 scheduler first determines the selected service interval (SI). Then an 802.11 e beacon interval is divided into an integer number of SIs, and stations 1204, 1206, 1208, 1210 are polled sequentially during each selected SI. Lastly, the access point 1202 scheduler computes the corresponding HCCA transmission opportunity (HCCA-TXOP) values for different stations by using their Q0S requests in TSPECs and allocates them to those QSTAs.

While aspects of the invention has been described in the context of a wireless personal area network wherein a plurality of nodes operate in the 60 GHz band, those skilled in the will appreciate that its teachings are applicable to communication networks operating in other frequencies.

Furthermore, it will be understood that, although antenna elements according to the embodiments of the present invention are described as being 60° or 90° angularly sectorised antennas, the present invention is not limited to such configurations. For example, antennas according to present invention may utilise one or more antenna elements to beamforrn towards a target node, such that the transmission patterns are not necessarily fixed or rigidly determined.

Moreover, it will be appreciated that the generic MAC protocol described herein is flexible in that various directional MAC techniques can be applied here according to the trade-off between performance improvement and implementation complexity. For example, in the above description it is assumed that directional transmission has the same range as omni-transmission. In fact, directional transmission usually has a higher range and this can be utilized in neighbour discovery to reach nodes via fewer but longer links, hereby saving network resources.

The skilled person will recognise that the above-described apparatus and method may be embodied as processor control code, for example on a carrier medium such as a disk, CD-or DVD-ROM, programmed memory such as read only memory (Firmware), or on a data carrier such as a n optical or electrical signal carrier. For many applications embodiments of the invention will be implemented on a DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array). Thus the code may comprise conventional programme code or microcode or, for example code for setting up pr controlling an ASIC or FPGA. The code may also comprise code for dynamically configuring re-configurable apparatus such as re-programmable logic gate arrays. Similarly the code may comprise code for a hardware description language. As the skilled person will appreciate, the code may be distributed between a plurality of coupled components in communication with one another. Where appropriate, the embodiments may also be implemented using code running on a field- (re)programmable analogue array or similar device in order to configure analogue hardware.

Claims (20)

1. CLAIMS: 1. A transceiver for use in a wireless communication network, the transceiver comprising: means for defining one or more access periods and respective subsequent data transmission periods such that said periods are synchronised with those of other nodes in the network; means for transmitting and receiving communication packets in any one or more of a plurality of directions, said means being adapted to transmit and receive control packets in all available directions during an access period; and means for contending for network medium access during an access period, said means for contending being operable to determine a likelihood of interference between a directional data packet communication in which the transceiver intends to participate in during a subsequent data transmission period and other intended or ongoing directional data packet communications in the network, wherein said likelihood is determined in consideration of any control packets received by the transceiver.
2. 2. A transceiver according to claim 1, further comprising means for storing information about nodes in the network, said information being indicative Qf directions to or locations of the nodes relative to the transceiver.
3. 3. A transceiver according to claim 1 or 2, wherein said likelihood is determined based on an estimated direction of arrival of said any received control packets.
4. 4. A transceiver according to any one of the preceding claims, wherein said likelihood is determined with reference to a common reference direction available to the transceiver.
5. 5. A transceiver according to any one of the preceding claims, wherein said likelihood is determined based on information contained in said any received control packets.
6. 6. A transceiver according to claim 5, wherein said information comprises at least one of: information identifying one or more nodes and information indicative of a direction of a corresponding intended directional data packet communication.
7. 7. A transceiver according to any one of the preceding claims, wherein the means for defming said periods is further operable to define a random wait period during an access period, the transceiver being prevented, before the expiry of said wait period, from iransmitting a control packet conveying a request to transmit directional data.
8. 8. A transceiver according to any one of the preceding claims, wherein the duration of an access period is dynamically adapted according to the network conditions.
9. 9. A transceiver according to any one of the preceding claims, wherein the access period comprises an Ad-hoc Traffic Indication Message (ATIM) window.
10. 10. A method of contending for medium access at a transceiver for use in a wireless communication network, the method comprising: defining one or more access periods and respective subsequent data transmission periods such that said periods are synchronised with those of other nodes in the network, the transceiver being adapted to transmit and receive control packets in all available directions during said access period; and determining, in consideration of any control packets received by the transceiver, a likelihood of interference between a directional data packet communication in which the transceiver intends to participate in during a data transmission period and other intended or ongoing directional data packet communications in the network.
11. 11. A method according to claim 10, and storing information about said nodes in the network, said information being indicative of directions to or locations of the nodes relative to the transceiver.
12. 12. A method according to claim 10 or 11, wherein said determining comprises estimating a direction of arrival of said received control packets.
13. 13. A method according to any one of claims 10 to 12, wherein said determining comprises referring to a common reference direction available to the transceiver.
14. 14. A method according to any one of claims 10 to 13, wherein said determining comprises extracting information contained in said any received control packets.
15. 15. A method according to any one of claims 10 to 14, further comprising defining a random wait period during an access period, the transceiver being prevented, before the expiry of said wait period, from transmitting a control packet conveying a request to transmit directional data.
16. 16. A method according to any one of claims 10 to 15, and dynamically adapting the duration of an access period according to the network conditions.
17. 17. A wireless communication network comprising a plurality of transceivers each according to any one of claims 1 to 9.
18. 18. A method of communicating in a wireless communication network having a plurality of transceivers, the method comprising: contending for medium access at each transceiver intending to participate in a directional data communication according to any one of claims 10 to 16; and implementing said intended participation at each transceiver determining that the likelihood of interference between a directional data packet communication in which the transceiver intends to participate in and other intended or ongoing directional data packet communications in the network is low, and in accordance with a successful control packet exchange.
19. 19. A computer program for implementing an apparatus according to any one of claims 1 to 9.
20. 20. A carrier medium carrying processor executable code for controlling a processor to carry out the method of any one of claims 10 to 16.