Joint Use Of Multi-Packet Reception And Network Coding For Performance Improvement

Abstract

Network coding and multiple packet reception (MPR) are used together in a wireless network. In at least one implementation, a novel medium access control (MAC) protocol is provided that enhances throughput in a wireless mesh network that uses network coding and MPR by providing fairness to information flows, rather than fairness to individual nodes.

Description

FIELD

This application relates generally to wireless communication and, more particularly, to techniques for enhancing throughput in a wireless network.

BACKGROUND

In a wireless system, bandwidth is typically a limited and expensive resource. Therefore, there is a general desire to develop communication strategies that use bandwidth efficiently. In a network scenario, this desire to use bandwidth efficiently may be realized by developing techniques for increasing throughput in the network.

SUMMARY

In accordance with one aspect of the concepts, systems, circuits, and techniques described herein, a method for use in a wireless mesh network that utilizes network coding and multi-packet reception (MPR) to distribute data in the network includes allocating transmission resources to nodes in the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network. In one embodiment, the method further includes limiting an amount of self-generated traffic that a relay node can transmit during its time slot allocation. The amount of self-generated traffic that the relay node can transmit may be limited, for example, to an average per node non-self-generated traffic level in the wireless mesh network. The amount of non-self-generated traffic to flow through the nodes of the network may be determined in one embodiment based on a current network topology and an amount of data stored in transmit buffers of nodes in the wireless mesh network. The allocation of transmission resources to nodes may be performed in a manner that provides fairness to information flows within the network.

In accordance with another aspect of the concepts, systems, circuits, and techniques described herein, a device for use in a wireless network that utilizes network coding and multi-packet reception (MPR) comprises: one or more wireless transceivers having MPR capability; a network coding module to perform network coding and/or network decoding for the device; and a resource allocation unit to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network. In one embodiment, the resource allocation unit may be configured to allocate transmission resources to nodes of the wireless network based, at least in part, on any one, or a combination of a current topology of the wireless network, a type of traffic to flow through the network, node types in the network, and an MPR capability of the receivers used in the network. The type of traffic may be determined in whole or in part by the number of destination nodes to receive the traffic. The type of traffic may also include information about priorities associated with the nodes of the network.

In accordance with still another aspect of the concepts, systems, circuits, and techniques described herein, a method for use in a wireless network that utilizes network coding and multi-packet reception (MPR) to distribute data in the network comprises allocating transmission resources to nodes of the wireless network based, at least in part, on a current topology of the wireless network. The allocation of transmission resources may be performed in a manner that provides fairness to information flows in the network, rather than fairness to individual nodes. In other embodiments, other fairness metrics may be used including for example, fairness to nodes, or priorities given to certain nodes. In addition, the allocation of transmission resources may be performed based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network at network saturation. In some embodiments, an amount of self-generated traffic that a relay node can transmit within its allocated resources may be limited. For example, in one embodiment, the self-generated traffic is limited to an average per node non-self-generated traffic level in the wireless network.

In accordance with a further aspect of the concepts, systems, circuits, and techniques described herein, a device for use in a wireless network that utilizes network coding and multi-packet reception (MPR) comprises: a wireless transceiver having MPR capability, a network coding module to perform network coding and/or network decoding for the device; and a resource allocation unit to allocate transmission resources to nodes of the wireless network based, at least in part, on a current topology of the wireless mesh network. In one embodiment, the resource allocation unit is configured to allocate transmission resources to nodes of the wireless network based, at least in part, on the topology of the wireless network and the type of traffic to flow through the network. In another embodiment, the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on the topology of the wireless network, the type of traffic to flow through the network, and an MPR capability of the receivers to be used in the network. In still another embodiment, the resource allocation unit is configured to allocate transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, and the node type of nodes in the network. In yet another embodiment, the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, the node type of each node in the network, and an MPR capability of the receivers to be used in the network. In still another embodiment, the resource allocation unit is to allocate transmission resources to nodes of the wireless network in a manner that provides fairness to information flows in the network. In another embodiment, the resource allocation unit is configured to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network.

In accordance with a still further aspect of the concepts, systems, circuits, and techniques described herein, a wireless network comprises: (a) a plurality of nodes that each include: (i) one or more wireless transceivers having multi-packet reception (MPR) capability; and (ii) a network coding module to perform network coding and/or network decoding for the node; and (b) resource allocation logic to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the wireless network. The resource allocation logic may be a centralized unit at a single location within the network or a system that is distributed throughout the network. In some implementations, the network may also have one or more nodes that do not have MPR capability.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram illustrating a simple network and showing how network coding may be used to enhance network throughput;

FIGS. 2, 3, 4, and 5 are schematic diagrams illustrating various network topologies that may exist within wireless networks;

FIG. 6 is a table illustrating calculated maximum network throughput gains for the topologies described in FIGS. 2, 3, 4, and 5 for various combinations of network coding and multi-packet reception (MPR) using a novel medium access control (MAC) protocol in accordance with an implementation;

FIG. 7 is a block diagram illustrating a resource allocation unit that may be used to provide resource allocation services from a single location in a network in accordance with an implementation;

FIG. 8 is a flowchart illustrating a method for allocating resources within a wireless network in accordance with an implementation; and

FIG. 9 is a block diagram illustrating an example wireless device architecture that may be used by a node in a wireless network in accordance with an embodiment.

DETAILED DESCRIPTION

A wireless mesh or ad-hoc network is a decentralized type of network that includes a number of wireless nodes that can intercommunicate through peer-to-peer wireless links. This is in contrast to an infrastructure-type wireless network where wireless nodes within a region communicate with one another, and/or with a larger network, through an associated wireless access point, base station, or other centralized control station. A wireless mesh network is a “fully connected” network if each of the nodes of the network can communicate with each of the other nodes through a direct wireless link. In a wireless mesh network that is not “fully connected,” one or more of the nodes in the network may serve as a relay node to relay messages between other nodes. A wireless mesh network may include any number of nodes (i.e., N≧2, where N is the number of nodes in the network). In addition, each of the wireless nodes in a wireless mesh network may include one or more wireless transceivers to support wireless communication with one or more of the other nodes in the network.

The route that a packet takes between a source node and a destination node in a wireless mesh network is known as a “path” through the network. A “one hop” path is a direct link between a source node and a destination node. A “two hop” path involves one link from a source node to a relay node and another link from the relay node to the destination node. Multi-hop paths that use multiple relay nodes between a source node and a destination node (e.g., a three hop path, etc.) may also be used in some networks. It should be appreciated that wireless mesh networks and ad-hoc networks may, in some implementations, include one or more nodes that perform an infrastructure type function. For example, mesh networks may, in some cases, include one or more mesh routers or mesh gateways to provide communication between mesh nodes and other networks (e.g., other mesh networks, the Internet, a private enterprise network, the Internet, a public switched telephone network (PSTN), a local area network, a municipal area network (MAN), a wide area network, and/or others).

As used herein, a “unicast” transmission is a transmission of a packet or packets from a specific source node to a specific destination node. A “broadcast” transmission is a transmission of a packet or packets from a source node to all other nodes in a network or sub-network. A “multicast” transmission is a transmission of a packet or packets from a specific source node to multiple, but not necessarily all, other nodes in the network. A sequence of packets traveling between a source node and a destination node or nodes (whether unicast, multicast, or broadcast) may be referred to as a “packet flow” or simply as a “flow.”

In a typical operational scenario within a mesh network, there may be a subset of nodes in the network that have packets ready for transmission (e.g., stored within a transmit buffer, etc.) at a particular point in time. For example, there may be J nodes in the network (J≦N) that each have k packets ready for transmission. The packets that are ready for transmission may include unicast traffic, multicast traffic, and/or broadcast traffic. The goal may be to successfully transfer all of these packets to their respective destinations. It is desirable that these transfers be made in a timely manner. One performance metric that is often discussed in connection with communication networks is “throughput,” which may be defined as an average rate of successful message delivery. Throughput may be specified for a particular communication channel or link, for traffic flowing through a particular node, or for an entire network or system (in which case, it may alternatively be referred to as “system throughput” or “aggregate throughput”). Throughput may be specified using any of a number of different formats including, for example, bits per second, data packets per second, data packets per time slot, and/or other formats. It is generally desirable to improve the throughput level for a channel, node, network, or system to make better use of available resources.

One technique that may be used to improve throughout in a wireless mesh network is network coding. In a network that does not use network coding, a relay node will typically just re-transmit received packets in the same form that they were received to facilitate transfer of the packets to their intended destinations. When network coding is used, on the other hand, a relay node may linearly combine data from multiple received packets to form a new packet (i.e., a coded packet) and then transmit the new packet in the network to facilitate the transfer of the multiple received packets to their intended destinations within a single transmission. Nodes receiving the coded packet may then extract relevant data packets from the coded packet using one or more decoding techniques. It has been shown that the use of network coding can significantly increase throughput in a mesh network.

FIG. 1 is a schematic diagram of a simple mesh network 30 that shows how network coding may be used to enhance network throughput. FIG. 1 illustrates a form of network coding known as COPE. COPE, and some other forms of network coding, rely on the broadcast nature of a wireless channel to allow nodes to overbear transmissions of neighbor nodes for eventual use in extracting data from coded packets received in the network. As shown in FIG. 1, the network 30 includes a first node 32 (node A), a second node 34 (node B), and a third node 36 (node R). First node 32 has a first packet 38 (i.e., packet a) that it wants to transfer to second node 34. Likewise, second node 34 has a second packet 40 (i.e., packet b) that it wants to transfer to first node 32. First and second nodes 32, 34 do not have a direct wireless link between them, so third node 36 is used as a relay node.

During a first time slot, first node 32 may transmit first packet 38 to third node 36, which stores the packet in memory. During a next time slot, second node 34 may transmit second packet 40 to third node 36. Third node 36 may then linearly combine first packet 38 and second packet 40 using an exclusive-OR (XOR) operation to generate a third packet 42 (i.e., the coded packet). Third node 36 then transmits (e.g., broadcasts) third packet 42, which is subsequently received by first node 32 and second node 34. Because first node 32 has knowledge of first packet 38, it is able to extract second packet 40 from the coded packet 42 (i.e., first node 32 is able to decode the coded packet 42). Likewise, because second node 34 has knowledge of second packet 40, it is able to extract first packet 38 from the coded packet 42.

If network coding was not being used, third node 36 would have to forward first packet 38 and second packet 40 during separate time slots. Using COPE, however, first packet 38 and second packet 40 could both be forwarded during a single time slot, thereby increasing network throughput. It should be appreciated that COPE is a relatively simple form of network coding that uses the XOR function to linearly combine packets being forwarded in a network. Other more complex forms of network coding also exist (e.g., randomized linear network coding that provides a linear combination of packets using randomly selected coefficients, etc.). In some embodiments described herein, a generalization of COPE may be used in conjunction with MPR and tailored resource allocation.

Another technique that may be used to enhance throughput in a wireless network is multi-packet reception (MPR). In traditional mesh networks, wireless devices were typically limited to receipt of one packet from one source within a particular time slot. Signals transmitted from other sources during the time slot were considered undesired interference and would often cause a “collision” to occur in the channel that could compromise the receiver's ability to detect and decode the desired signal. MPR is a technique that allows a receiver to receive packets from multiple different sources simultaneously, thereby reducing the negative impact of collisions in the channel. Various different wireless transmission technologies may be utilized to support the implementation of MPR in a wireless network including, for example, code division multiple access (CDMA), frequency hopping spread spectrum, direct sequence spread spectrum, multiple input/multiple output (MIMO), orthogonal frequency division multiple access (OFDMA), and/or others. In one approach, which will be referred to herein as heterogeneous MPR, MPR may be implemented using multiple different radio technologies. This may include, for example, multiple radios operating in accordance with different wireless standards (e.g., two or more of IEEE 802.11, IEEE 802.15, IEEE 802.16, Bluetooth, Zigbee, Ultrawideband, third generation mobile communication standards, fourth generation mobile communication standards, satellite communications standards, wireless cellular standards, and/or others). MPR can improve throughput in a wireless network by, among other things, relieving channel contention and multi-user interference issues, reducing data loss due to collisions, and increasing the amount of data that can be transferred per unit of time. In many cases, MPR-enabled receivers are needed for MPR to be successfully implemented within a network. The number of packets that an MPR-enabled receiver is capable of receiving simultaneously will be referred to herein as the MPR coefficient (m) of the receiver.

As will be described in greater detail, in some implementations described herein, network coding and MPR are used together within a wireless network to provide an enhanced level of performance (e.g., increased system throughput, reduced delay, etc.) in the network. For example, referring back to the simple mesh network 30 of FIG. 1, MPR techniques may be used to improve throughput in this network. That is, MPR may be combined with COPE by allowing first and second nodes 32, 34 to transmit first and second packets 38, 40 at the same time (i.e., during the same time slot). Third node 36 may then transmit third (coded) packet 42 during a subsequent time slot. As will be appreciated, the data transfer operation of delivering first packet 38 to second node 34 and second packet 40 to first node 32 is performed in less time (i.e., greater throughput) using this approach than using COPE alone. As will be described in greater detail, this concept may be extended to other network topologies and scenarios.

In addition to the above, in some implementations, unique medium access control (MAC) techniques are provided that are capable of further enhancing network performance. In one approach, for example, a novel MAC protocol is provided that can further enhance throughput in a network that uses both network coding and MPR. The MAC protocol allocates transmission resources in a network in a manner that provides fairness to information flows, rather than fairness to individual nodes as specified in, for example, the IEEE 802.11 wireless networking standard. The current IEEE 802.11 MAC protocol allocates the same amount of network resources to bottlenecked nodes as to edge nodes, despite the former's need to use some of these resources for relaying. It has been found that the improved MAC techniques described herein, in combination with network coding, MPR, and fairness to flows are capable of increasing achievable throughput by as much as 6.3 times or more over networks that do not use network coding and MPR.

It has been found that the overall improvement in throughput that can be achieved in a wireless mesh network that is using both MPR and network coding depends upon a number of different factors. These different factors may include, for example, the particular network topology being used, the MPR capability of the network, the type of traffic being carried in the network, and/or other factors. In some implementations described herein, factors such as these are taken into consideration during an allocation of transmission resources (e.g., time slots, frequency channels, CDMA codes, polarizations, OFDMA subchannels, and/or others, including combinations of the above) to nodes within a mesh network.

In the discussion that follows, a particular network model will be assumed for purposes of analysis. In this model, packets are never delayed. That is, if a node in the network has only a single codable packet, it will not wait for another packet to arrive before transmitting a signal. Rather, the node will transmit the packet uncoded at the first opportunity. The model also assumes that all packets in the network are of the same length. Third, packets headed towards the same next hop will never be coded together under the model. This is because, if this were not required, the node associated with the next hop would not have enough information available to decode the coded packet (i.e., the node would not have the necessary degrees of freedom to decode), since fewer coded packets are transmitted than original packets.

In the model, it will be assumed that each node can randomly generate packets and each packet may then be transmitted through a relay node to a destination. The relay node will be assumed to be fully connected regardless of the network topology and packets generated at the relay node will require only a single hop to reach their intended destination. A unicast transmission will be considered complete when all packets from each source node successfully reach their destinations. A broadcast transmission will be considered complete when all nodes within the network have received each packet from all of the sources. Furthermore, under the model, it will be assumed that each node is half-duplex and, as a result, a node cannot receive another node's transmissions while it is transmitting.

Under the model, it will also be assumed that each node can receive multiple simultaneous packets without delay or loss. In addition, if a node is not transmitting and it has direct communication or can overhear another node, it will automatically receive any transmission made by that node and will be able to use that information to decode any coded messages it receives.

FIGS. 3, 4, 5, and 6, are schematic diagrams illustrating various network topologies that may exist within wireless mesh networks. As will be described in greater detail, it has been found that optimal techniques for implementing MPR and network coding within a wireless mesh network may depend upon a current network topology of the network. The network topologies of FIGS. 3, 4, 5, and 6 are topologies that naturally form bottlenecks and create congestion. In the discussion that follows, each of the network topologies of FIGS. 3, 4, 5, and 6 will be briefly described. Enhanced techniques for managing network traffic in networks using these and other topologies will then be discussed.

FIG. 2 is a diagram illustrating a cross topology 50 having fives nodes 52, 54, 56, 58, 60. The first four nodes 52, 54, 56, 58 are edge nodes and the fifth node 60 is a central relay node. Each of the nodes 52, 54, 56, 58, 60 of the cross topology 50 can directly transmit and receive information to/from every other node. The only exception is that each edge node is not connected with an edge node on an opposite side of relay node 60 (e.g., node 52 is not directly connected to node 56, and node 54 is not directly connected to node 58). FIG. 3 is a diagram illustrating a modified cross topology 70 having fives nodes 72, 74, 76, 78, 80. The modified cross topology 70 is similar to the cross topology 50 of FIG. 2, except edge node 76 and edge node 78 are not directly connected and therefore cannot overhear each other's transmissions.

FIG. 4 is a diagram illustrating an X topology 90 having five nodes 92, 94, 96, 98, 100. The first four nodes 92, 94, 96, 98 are edge nodes and the fifth node 100 is a central relay node. In the X topology 90, it is assumed that nodes 92, 94, 100 are directly connected to one another and nodes 96, 98, and 100 are directly connected to one another, but there is no direct connection between nodes 92, 94 in a first edge node group X₁ and nodes 96, 98 in a second edge node group X₂. That is, all traffic between a node in group X₁ and a node in group X₂ must take place through relay node 100. FIG. 5 is a diagram illustrating a modified X topology 90 having fives nodes 112, 114, 116, 118, and 120. Modified X-topology 110 is similar to the X topology 90 of FIG. 4, except that there is no direct connection between nodes 116 and 118.

In analyzing the various network topologies set out above, it will be assumed that each edge node transmits all of its available packets to the corresponding relay node. Once every node has sent all of its available packets to the relay node, the relay node will either identify coding opportunities and transmit a set of coded messages optimized for the network topology being used or send the packets uncoded. When MPR is being used, m packets will be allowed to be sent from different sources in a single time slot. Since MPR, in the context of the topologies analyzed, is a method of avoiding collisions due to hidden nodes, the existing carrier sense multiple access with collision avoidance (CSMA/CA) protocols of the IEEE 802.11 standard will be followed for each m=2 case. For cases involving m=4, an extended version of CSMA/CA will be used to allow each edge node to transmit in the same time slot to the relay node. In the analysis, the effects of collisions due to either hidden nodes or identical back off times will not be considered because the effects on total throughput are small in relation to the effects of fairness provided by the 802.11 MAC. In addition, the potential gains resulting from MPR alleviating impacts of the exposed terminal problem caused by IEEE 802.11 virtual CS mechanisms will not be considered.

As described above, it has been found that, when implementing network coding with MPR, the medium access control (MAC) protocol can impact overall performance in a significant manner. For networks with bottlenecks, such as networks using the network topologies of FIGS. 3, 4, 5, and 6, the parameters used to ensure fairness among competing nodes at saturation are critical to ensuring that enhanced throughput is achieved. In the current IEEE 802.11 MAC, time slots are distributed equally among all competing nodes within a network, regardless of topology. As network load increases, therefore, this MAC will limit each edge node's traffic to the relay node, while the rate of traffic introduced to the network by the relay node will not be similarly constrained. In each of the different topologies, therefore, nodes sending both relayed traffic and self-generated traffic will inherently send more of their own self-generated traffic and the effectiveness of opportunistic network coding will be reduced. In at least one aspect of the techniques and concepts described herein, a novel MAC protocol is provided that allocates time slots (or other transmission resources) to competing nodes using a basic knowledge of the current network topology and the MPR capabilities of each receiver. In this manner, the MAC is able to provide fairness to information flows, rather than individual nodes.

In one approach, the novel MAC allocates transmission resources to nodes in a manner that is based, at least in part, on an amount of non-self-generated traffic to flow through each node. The amount of non-self-generated traffic to flow through a node may be based upon, for example, an amount of data stored in transmit buffers of the nodes of the network and the network topology. Using this approach, a relay node will typically be allocated more resources than an edge node because it must also relay information. In addition, in some implementations, the MAC may require each node relaying information to limit an amount of self-generated traffic to the average per node non-self-generated traffic being relayed. While allocating fewer resources to flows originating at the relay node and more resources to flows originating at the edge nodes yields even higher throughputs, the MAC ensures that each flow of information is given the same priority. In the discussion that follows, techniques are described for allocating resources (e.g., time slots and/or other transmission resources) to the nodes of networks based, at least in part, on network topology and MPR coefficient in accordance with an implementation. The discussion will be made with reference to the network topology diagrams of FIGS. 3, 4, 5, and 6.

For the cross topology 50 of FIG. 2, the allocation of resources is the same for both unicast and broadcast transmission. This is assuming that that there is no constraint on the order in which each node transmits. When network coding is not used, relay node 60 requires a number of time slots equal to the sum of source nodes, N. When network coding is used, throughput can be maximized by ensuring that relay node 60 codes a maximum number of uncoded packets together. Using MPR in this scenario can potentially prevent each node from immediately decoding any coded message sent by relay node 60 since we are allowing nodes with indirect lines of communication to transmit at the same time. For example, when m=2, relay node 60 needs to send two coded packets, each combined in a different manner, to ensure that each edge node has the necessary degrees of freedom to decode the packet. Generalizing for N and m gives the following:

$s_i=\{\begin{array}{cc}\frac{1}{\lceil \left(N-1\right)/m\rceil +N}& \mathrm{without}\mathrm{NC}\\ \frac{1}{\lceil \left(N-1\right)/m\rceil +m_c+1}& \mathrm{with}\mathrm{NC}\end{array}s_R=\{\begin{array}{cc}\frac{N}{\lceil \left(N-1\right)/m\rceil +N}& \mathrm{without}\mathrm{NC}\\ \frac{m_c+1}{\lceil \left(N-1\right)/m\rceil +m_c+1}& \mathrm{with}\mathrm{NC}\end{array}$

where s_i is the fraction of time slots allocated to each edge node 52, 54, 56, 58 and s_R is the fraction of time slots allocated to relay node 60. The variable m_c depends upon whether or not carrier sense multiple access (CSMA) is being used and may be defined as follows:

$m_c=\left\{\begin{array}{ccc}m& m=1\mathrm{or}2& \mathrm{CSMA}\mathrm{is}\mathrm{not}\mathrm{used}\\ m-1& m=2& \mathrm{CSMA}\mathrm{is}\mathrm{used}\\ m-1& m=4& \end{array}\right\}$

It should be appreciated that, when CSMA is used, only nodes on opposite sides of relay node 60 are allowed to transmit in the same time slot. Enforcement of this limitation can result in a significant gain in throughput for small N, but the effect may become less significant as N grows.

For unicast traffic in the X topology 90 of FIG. 4, the maximum number of packets that can be coded together is two and only packets from different sets can be usefully coded together. As a result, the fraction s_i^U of time slots allocated for unicast traffic is:

$s_i^U=\{\begin{array}{cc}\frac{1}{\lceil \left(N-1\right)/m\rceil +N}& \mathrm{without}\mathrm{NC}\\ \frac{1}{\lceil \left(N-1\right)/m\rceil +\max \left(X_1,X_2\right)+1}& \mathrm{with}\mathrm{NC}\end{array}s_R^U=\{\begin{array}{cc}\frac{N}{\lceil \left(N-1\right)/m\rceil +N}& \mathrm{without}\mathrm{NC}\\ \frac{\max \left(X_1,X_2\right)+1}{\lceil \left(N-1\right)/m\rceil +\max \left(X_1,X_2\right)+1}& \mathrm{with}\mathrm{NC}\end{array}$

where |X₁| is a number of nodes in a first edge node group and |X₂| is a number of nodes in a second edge node group. When the number of nodes, packets, and/or destinations on either side of relay node 100 is asymmetric, the number of packets that will be coded together is equal to the minimum of the number of packets originating from set X₁ or X₂. The remaining packets will be forwarded uncoded.

For broadcast traffic in the X topology 90 of FIG. 4, the equations set out above for unicast traffic in the X topology still apply when network coding is not used. When network coding is allowed, there is a possibility that each destination node will require a maximum of one additional degree of freedom per node for m=2 or three additional degrees of freedom per node for m=4 when either |X₁|≧m or |X₂|≧m and the order of node transmission is not enforced. Providing these additional degrees of freedom can be accomplished by relay node 100 sending at most three coded packets, where the sum of all of the native edge node packets are included in each coded transmission and each coded packet is combined in a different manner. Each edge node's fraction of time slots is maximized when the cardinality of each set of nodes, X₁ and X₂, is equal. The fraction of time slots is minimized when the cardinality of each set is asymmetric; and for m>1, transmission from edge nodes 92, 94, 96, 98 to relay node 100 is asymmetric (i.e., multiple nodes from a single set transmit at the same time). The fraction s_i^B of time slots for each edge node when network coding is used in the X topology for broadcast traffic may be expressed as follows:

$\frac{1}{\lceil \left(N-1\right)/m\rceil +\max \left(X_1,X_2\right)+m}\le s_i^B\le s_i^U$

where s_i^U is the fraction of time slots of an edge node for unicast traffic. Similarly, the fraction of time slots for the relay node when network coding is used for broadcast traffic may be expressed as follows:

$s_R^U\le s_R^B\le \frac{\max \left(X_1,X_2\right)+m}{\lceil \left(N-1\right)/m\rceil +\max \left(X_1,X_2\right)+m}$

where s_R^U is the fraction of time slots of a relay node for unicast traffic.

For the partial cross topology 70 of FIG. 3 and the partial X topology 110 of FIG. 5, the fraction of time slots used by each node of the network will be similar to those described above for the full cross topology 50 of FIG. 2 and the full X topology 90 of FIG. 4. As discussed above, the use of the described resource allocation scheme has been shown to result in significant increases in network throughput performance (at saturation) in networks implementing both network coding and MPR, as well as in networks that can use only network coding or only MPR. The scheme can also provide throughput gains in networks that use routing. FIG. 6 is a diagram illustrating a table 130 showing maximum network throughput gains for the various topologies described above for various combinations of network coding and MPR using the new MAC described above. As shown, when using both network coding and MPR, gains in maximum throughput of up to 6.25 may be achieved over implementations using the current IEEE 802.11 MAC protocol. These techniques can be extended for use in larger networks and, it is expected, may provide even further gains in throughput as network size increases.

FIG. 7 is a block diagram illustrating a resource allocation unit 150 that may be located within a node device or other device to provide resource allocation services for mesh networks that use both network coding and MPR in accordance with an implementation. Resource allocation unit 150 may be used to allocate resources within an ad-hoc or mesh wireless network in accordance with principles discussed herein. Resource allocation unit 150 may be part of, for example, a medium access control layer of a wireless node device. As illustrated, resource allocation unit 150 may receive inputs for various parameters of a wireless network (or actively solicit, poll, discover, or retrieve parameter information) and use these parameter values to determine how to allocate resources to one or more nodes of the network. As shown, in one embodiment, the parameter values may include, for example, network topology, traffic type, node type, and MPR capability. Other parameters values or combinations of parameter values may be used in other implementations.

The traffic type may include, for example, whether the traffic is unicast, multicast, or broadcast traffic. In some embodiments, the traffic type may also include a “priority” of the traffic. That is, in some implementations, different source nodes (or users associated with those nodes) may have different priorities based on, for example, the importance of the node/user/message, delay constraints associated with the node/user/message, quality-of-service (QoS) associated with the node/user/message, and/or other factors. This information may be taken into consideration when making the resource allocation decision. The “node type” may include, for example, whether a node is an edge node or a relay node.

The MPR capability that is used to perform resource allocation will typically depend on the MPR capabilities of the nodes in a network. In different implementations, the nodes in the network may all have the same MPR capability or different nodes may have different MPR capability. If all nodes have the same MPR capability, then that MPR capability of the nodes may be used to perform resource allocation. If different nodes have different capabilities (e.g., some have a higher MPR coefficient and some have a lower MPR coefficient), then the MPR capability used to perform resource allocation may be determined in a number of different ways. In one approach, the lower MPR coefficient among the nodes may be selected. In another approach, the higher MPR coefficient among the nodes may be selected. In still another approach, the higher MPR coefficient may be used by the transmitting nodes and the lower MPR coefficient may be used to perform backfilling and to transmit to the nodes having a lower MPR coefficient. In yet another approach, the transmitting nodes may use the lower MPR coefficient and backfilling operations may use the higher MPR coefficient.

In at least one implementation, resource allocation unit 150 may determine network allocations in a manner that provides fairness to data flows, rather than fairness to individual nodes. For example, resource allocation unit 150 may allocate resources to nodes in a manner that is proportional to an amount of non-self-generated traffic flowing through each node when the network saturates. Resource allocation unit 150 may also require each node relaying information to limit an amount of self-generated traffic to an average per node non-self-generated traffic being relayed. In some implementations, resource allocation unit 150 may use one or more of the equations set out herein (and/or other equations) to determine a resource allocation(s). In some other implementations, resource allocation unit 150 may determine resource allocations based on one or more additional or alternative metrics such as, for example, priorities associated with nodes/users/messages, message delay constraints, and/or others. An MPR coefficient of a wireless device will typically be a fixed value that is based on the capabilities of the device. Network topology and node type information may be available from, for example, topology discovery and/or route discovery protocols that are already active in a network.

In one possible implementation, when a mesh network is originally formed, one of the nodes of the network may be chosen to perform a resource allocation function for the network. The selected node may then activate a corresponding resource allocation unit 150 and begin to collect information necessary for determining resource allocations for the network. In some other implementations, a distributed approach may be used where each node may determine its own resource allocation using a corresponding resource allocation unit 150, although some coordination between nodes may be needed to provide a workable allocation for the overall network using this approach. In at least one implementation, resource allocation unit 150 may be implemented within a MAC module associated with a wireless transceiver or a host processor.

FIG. 8 is a flow diagram showing a process for allocating resources within a wireless mesh network that uses both network coding and MPR.

The rectangular elements (typified by element 172 in FIG. 8) are herein denoted “processing blocks” and may represent computer software instructions or groups of instructions. It should be noted that the flow diagram of FIG. 8 represents one exemplary embodiment of the design described herein and variations in such a diagram, which generally follow the process outlined, are considered to be within the scope of the concepts, systems and techniques described and claimed herein.

Alternatively, the processing blocks may represent operations performed by functionally equivalent circuits such as a digital signal processor circuit, an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA). Some processing blocks may be manually performed while other processing blocks may be performed by a processor. The flow diagram does not depict the syntax of any particular programming language. Rather, the flow diagram illustrates the functional information one of ordinary skill in the art requires to fabricate circuits and/or to generate computer software to perform the processing required of the particular apparatus. It should be noted that many routine program elements, such as initialization of loops and variables and the use of temporary variables are not shown. It will be appreciated by those of ordinary skill in the art that unless otherwise indicated herein, the particular sequence described is illustrative only and can be varied without departing from the spirit of the concepts described and/or claimed herein. Thus, unless otherwise stated, the processes described below are unordered meaning that, when possible, the sequences shown in FIG. 8 can be performed in any convenient or desirable order.

Turning now to FIG. 8, a method 170 for allocating resources within a wireless mesh network that uses both network coding and MPR will be described. A network topology of the network may first be determined (block 172). The network topology may be one of the topologies described previously, or a different topology. A traffic type to be transmitted by one or more nodes (e.g., broadcast, unicast, etc.) may then be determined (block 174). The node type of each node of the network may next be determined (block 176). This node type may include, for example, whether a node is an edge node or a relay node. An MPR capability to be used in the network may also be determined (block 178). The MPR capability for the network may be determined based on the capabilities of the nodes of the network. A fairness policy of the network may also be determined (block 180). Once this information has been collected, resource allocations may be computed for the network using the network topology, the traffic type, the node type, the MPT capability, and/or the fairness policy information (block 182). As described previously, in some implementations, resources may be allocated to nodes of the network in a manner that provides fairness to flows rather than fairness to individual nodes. In one possible approach, for example, resources may be allocated proportional to an amount of non-self-generated traffic flowing through each node when the network saturates. Resources may also be allocated so that an amount of self-generated traffic being transmitted by each relay node is limited to an average per node non-self-generated traffic being relayed.

The method 170 may be performed within, for example, one of the nodes of a mesh network and the resulting time allocations may then be communicated to the other nodes. In another possible approach, the method 170 may be performed within each of the nodes of the network, with some possible inter-node coordination, to determine a time slot allocation for the node. In some implementations, the method 170 may be repeated continually in the network so that an optimal or near optimal resource allocation is maintained.

FIG. 9 is a block diagram illustrating an example wireless device architecture 200 that may be used for a node in a wireless mesh network in accordance with an embodiment. As illustrated, the wireless device architecture 200 may include: one or more digital processors 202, a memory 204, a wireless transceiver 206, and a network coding module 208. A bus 210 and/or one or more other transmission structures may be provided for establishing interconnections between various components of the architecture 200. The wireless transceiver 206 may be coupled to one or more antennas 212 and/or other transducers to facilitate the transmission and/or reception of wireless signals.

Digital processor(s) 202 may include one or more digital processing devices that are capable of executing programs to provide one or more functions and/or services to a user. Digital processor(s) 202 may be used to, for example, execute an operating system of a corresponding wireless device. Digital processor(s) 202 may also be used to, for example, execute user application programs. In addition, digital processor(s) 202 may be used to implement, either partially or fully, one or more of the processes or techniques described herein in some implementations. Digital processor(s) 202 may include any type of digital processing device including, for example, a general purpose microprocessor, a digital signal processor (DSP), a controller, a microcontroller, an application specific integrated circuits (ASIC), a field programmable gate array (FPGA), a programmable logic array (PLA), a programmable logic device (PLD), a reduced instruction set computer (RISC), and/or others, including combinations of the above.

Wireless transceiver 206 may include any type of transceiver that is capable of supporting wireless communication with one or more remote wireless entities. In various implementations, wireless transceiver 206 may be configured in accordance with one or more wireless networking standards and/or wireless cellular standards. In some implementations, multiple wireless transceivers may be provided to support operation with different networks or systems in a surrounding environment. As illustrated in FIG. 9, in some implementations, wireless transceiver 206 may include a medium access control (MAC) module 214 to facilitate medium access operations on an associated wireless network medium or channel. In some embodiments, some of the MAC functionality may also be provided in digital processor(s) 32 (e.g., MAC module 220). In addition, in some implementations, wireless transceiver 36 may include a multi-packet reception (MPR) module 46 to facilitate the use of MPR by a corresponding wireless device.

Memory 204 may include any type of structure that is capable of storing digital information. The digital information may include, for example, digital user data, computer executable instructions and/or programs, or any other type of data. Memory 204 may include, for example, magnetic data storage devices, disc based storage devices, optical storage devices, semiconductor memories, read only memories (ROMs), random access memories (RAMs), non-volatile memories, flash memories, USB drives, compact disc read only memories (CD-ROMs), DVDs, Blu-Ray disks, magneto-optical disks, erasable programmable ROMs (EPROMs), electrically erasable programmable ROMs (EEPROMs), magnetic or optical cards, and/or others.

Network coding module 208 is operative for performing network coding operations and/or network decoding operations for the mobile device. In some implementations, network coding module 208 may be called upon to, for example, generate coded packets for re-transmission by combining packets received from other mobile devices. In some implementations, network coding module 208 may combine packets using an exclusive-OR function, such as used in COPE. However, in other implementations, other or additional forms of network coding may be implemented within network coding module 208 such as, for example, random linear network coding (RLNC). Although illustrated as a separate unit in FIG. 9, it should be appreciated that, in some implementations, the network coding module may be implemented within digital processor(s) 202.

It should be appreciated that the mobile device architecture 200 of FIG. 9 represents one possible example of an architecture that may be used in a implementation. Other architectures may alternatively be used. It should also be appreciated that all or part of the various devices, processes, or methods described herein may be implemented using any combination of hardware, firmware, and/or software.

Although discussed above primarily in the context of wireless mesh networks, it should be appreciated that the techniques, devices, and systems described herein can be used in other types of wireless network.

Having described preferred embodiments which serve to illustrate various concepts, structures and techniques which are the subject of this patent, it will now become apparent to those of ordinary skill in the art that other embodiments incorporating these concepts, structures and techniques may be used. Accordingly, it is submitted that that scope of the patent should not be limited to the described embodiments but rather should be limited only by the spirit and scope of the following claims.

Claims

1. A method for use in a wireless network that utilizes network coding and multi-packet reception (MPR) to distribute data in the network, the method comprising: allocating transmission resources to nodes in the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network.

2. The method of claim 1, further comprising: limiting an amount of self-generated traffic that a relay node in the wireless network can transmit within its allocated transmission resources.

3. The method of claim 2, wherein: limiting an amount of self-generated traffic that a relay node can transmit within its allocated transmission resources includes limiting the amount of self-generated traffic to an average per node non-self-generated traffic level in the wireless network.

4. The method of claim 1, further comprising: determining, before allocating transmission resources, the amount of non-self-generated traffic to flow through the nodes of the wireless network based on a current network topology and an amount of data stored in transmit buffers of nodes in the wireless network.

5. The method of claim 1, wherein: allocating transmission resources to nodes in the wireless network includes allocating the transmission resources in a manner that provides fairness to information flows within the network.

6. The method of claim 1, wherein: the wireless network includes a wireless mesh network.

7. The method of claim 1, wherein the transmission resources include one or more of: time slots, frequency channels, spread spectrum codes, polarizations, and orthogonal frequency division multiple access (OFDMA) sub-channels.

8. A device for use in a wireless network that utilizes network coding and multi-packet reception (MPR), the device comprising: one or more wireless transceivers having MPR capability,a network coding module to perform network coding and/or network decoding for the device; anda resource allocation unit to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network.

9. The device of claim 8, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on a topology of the wireless network.

10. The device of claim 8, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on the topology of the wireless network and a type of traffic to flow through the network.

11. The device of claim 10, wherein: the type of traffic to flow through the network includes priorities associated with the traffic.

12. The device of claim 8, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on the topology of the wireless network, the type of traffic to flow through the network, and an MPR capability of the network.

13. The device of claim 8, wherein: the resource allocation unit is configured to allocate transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, and the node type of nodes in the network.

14. The device of claim 8, wherein: the wireless network includes a wireless mesh network.

15. A method for use in a wireless network that utilizes network coding and multi-packet reception (MPR) to distribute data in the network, the method comprising: allocating transmission resources to nodes of the wireless network based, at least in part, on a current topology of the wireless network.

16. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network and the type of traffic to flow through the network.

17. The method of claim 16, wherein: the type of traffic to flow though the network includes priorities associated with the traffic.

18. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, and an MPR capability of the network.

19. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, and the node type of nodes in the network.

20. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, the node type of nodes in the network, and an MPR capability of the network.

21. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network in a manner that provides fairness to information flows in the network.

22. The method of claim 15, wherein: allocating transmission resources includes allocating transmission resources to nodes of the wireless network in a manner that accounts for message priorities in the network.

23. The method of claim 15, wherein: allocating transmission resources includes allocating resources to nodes based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network.

24. The method of claim 23, further comprising: limiting an amount of self-generated traffic that a relay node can transmit within its allocated transmission resources.

25. The method of claim 24, wherein: limiting an amount of self-generated traffic that a relay node can transmit within its allocated transmission resources includes limiting the amount of self-generated traffic to an average per node non-self-generated traffic level in the wireless network.

26. The method of claim 15, wherein: allocating transmission resources includes allocating one or more of: time slots, frequency channels, spread spectrum codes, polarizations, and orthogonal frequency division multiple access (OFDMA) sub-channels.

27. The method of claim 15, wherein: the wireless network includes a wireless mesh network.

28. A device for use in a wireless network that utilizes network coding and multi-packet reception (MPR), the device comprising: a wireless transceiver having MPR capability;a network coding module to perform network coding and/or network decoding for the device; anda resource allocation unit to allocate transmission resources to nodes of the wireless network based, at least in part, on a current topology of the wireless mesh network.

29. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless mesh network based, at least in part, on the topology of the wireless network and the type of traffic to flow through the network.

30. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless mesh network based, at least in part, on the topology of the wireless network, the type of traffic to flow through the network, and an MPR capability of the network.

31. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless mesh network based, at least in part, on the current topology of the wireless network, the type of traffic to flow through the network, and the node type of nodes in the network.

32. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on the topology of the wireless mesh network, the type of traffic to flow through the network, the node type of nodes in the network, and an MPR capability of the network.

33. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network in a manner that provides fairness to information flows in the network.

34. The device of claim 28, wherein: the resource allocation unit is to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the network.

35. The device of claim 28, wherein: the device is part of one of the following: a network router, a portable computer, a smart phone, a satellite communicator, a satellite-based transceiver, a network storage device, a tablet, a mobile wireless apparatus, or a stationary wireless apparatus.

36. A wireless network comprising: a plurality of nodes that each include: one or more wireless transceivers having multi-packet reception (MPR) capability; anda network coding module to perform network coding and/or network decoding for the node; andresource allocation logic to allocate transmission resources to nodes of the wireless network based, at least in part, on an amount of non-self-generated traffic to flow through the nodes of the wireless network.

37. The wireless network of claim 36, further comprising: at least one additional node that does not have MPR capability.

38. The wireless network of claim 36, wherein: the wireless network includes a wireless mesh network.

39. The wireless network of claim 36, wherein: the resource allocation logic is distributed across multiple locations in the network.

40. The wireless network of claim 36, wherein: the resource allocation logic includes a centralized resource allocation unit located at a single location within the network.