This disclosure relates to a method and apparatus for providing full-duplex operation in a multipoint-to-multipoint network.
The background description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the inventors hereof, to the extent the work is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted to be prior art against the present disclosure.
Many full-duplex systems operate in a point-to-point mode in which two nodes communicate between themselves. In such a case, Medium Access Control (MAC) is not required given that each of the two nodes can always transmit while simultaneously receiving from the other node. Examples of full-duplex point-to-point systems are Ethernet 1000BASE-T and 10GBASE-T systems.
Some full-duplex systems operate in a point-to-multipoint mode in which one node (sometimes referred to as a “master node” or “access point”) transmits to multiple other nodes and receives transmission from the other nodes. There is no direct communication between the other nodes. The master node or access point transmits in the downstream direction, and the other nodes share the upstream transmissions. The master node or access point is thus the only node that can initiate a full-duplex transmission.
A method of conducting full-duplex transmission between first and second nodes in a communications system having more than two nodes includes the issuing of a start signal on a channel by the first node, wherein the start signal signals that full-duplex transmission is to begin, and identifies the second node as a node with which full-duplex communication is to occur. The method also includes, following the issuing of the start signal, beginning full-duplex transmission by the first node and by the second node.
In such a method, each node of the first and second nodes signals a respective duration of its respective transmission, and duration of the full-duplex transmission between the first and second nodes is determined by whichever respective duration is longer.
The method further includes the one of the first and second nodes whose respective transmission has a shorter duration transmitting additional symbols to extend its transmission to match the duration of the transmission of the other of the first and second nodes. Alternatively, the method includes the first node signaling a required transmission duration based on its data to be transmitted, and the second node adjusting its transmission to fit the required transmission duration. In such a method, the second node adjusting its transmission includes the second node shortening its transmission by decreasing the amount of data to be sent, or the second node adjusting its transmission includes the second node transmitting additional symbols to extend its transmission to match the duration of the transmission of first a node.
The method further includes performing echo cancellation on the full-duplex transmission to remove, from a transmission in one direction, an echo of a transmission in another direction.
In the method, the full-duplex transmission may be performed using orthogonal frequency division modulation (OFDM), the first node transmits on a first set of OFDM subcarriers, and the second node transmits on a second set of OFDM subcarriers.
In the method, the full-duplex transmission is performed using orthogonal frequency division modulation (OFDM), and at least one of the first node and the second node transmits a signal on at least one OFDM subcarrier to signal other nodes that the channel is in use.
A method of conducting full-duplex transmission between nodes in a communications system having more than two nodes includes establishing a medium access plan assigning respective temporal slots to respective pairs of nodes, and on opening of a respective temporal slot, initiating full-duplex transmission between nodes in the respective pair of nodes to which the respective temporal slot is assigned, wherein any node in the communications system is capable of full-duplex communication with any other node according to the medium access plan.
In the method, the initiating full-duplex transmission includes issuing of a start signal on a channel by one of the nodes in the respective pair of nodes, signaling that full-duplex transmission is to begin, and following the issuing of the start signal, beginning full-duplex transmission by both of the nodes in the respective pair of nodes.
In the method, each node in the respective pair of nodes signals a respective duration of its respective transmission, and duration of the full-duplex transmission between the nodes in the respective pair of nodes is determined by whichever respective duration is longer. The one of the nodes in the respective pair of nodes whose respective transmission has a shorter duration may transmit additional symbols to extend its transmission to match the duration of the transmission of the other of the first and second nodes.
The method also includes a first one of the nodes in the respective pair of nodes signaling a required transmission duration based on its data to be transmitted, and a second one of the nodes in the respective pair of nodes adjusting its transmission to fit the required transmission duration.
The method also includes performing echo cancellation on the full-duplex transmission to remove, from a transmission in one direction, an echo of a transmission in another direction.
In the method, the full-duplex transmission is performed using orthogonal frequency division modulation, a first set of subcarriers is used for transmission by a first node in the respective pair of nodes, and a second set of subcarriers is used for transmission by a second node in the respective pair of nodes.
A node, for use in a communications network including at least two nodes, is configured to initiate full-duplex communication with any other node by issuing a start signal on a channel, wherein the start signal signals that full-duplex transmission is to begin, and identifies a second node as a node with which full-duplex communication is to occur, and following the issuing of the start signal, beginning full-duplex communication. A communications system includes a plurality of that node.
In such a communications system, each node signals a respective duration of its respective transmission, and duration of the full-duplex transmission is determined by whichever respective duration is longer.
The communications system further includes the node whose respective transmission has shorter duration transmitting additional symbols to extend its transmission to match the duration of the transmission of the node whose respective transmission has a longer duration.
The communications system further includes one node signaling a required transmission duration based on its data to be transmitted, and the other node adjusting its transmission to fit the required transmission duration.
In the communications system, the full-duplex transmission is performed using orthogonal frequency division modulation (OFDM), one node transmits on a first set of OFDM subcarriers, and the other node transmits on a second set of OFDM subcarriers.
In the communications system, the full-duplex transmission is performed using orthogonal frequency division modulation (OFDM), and at least one of the nodes transmits a signal on at least one OFDM subcarrier to signal other nodes that the channel is in use.
Further features of the disclosure, its nature and various advantages, will be apparent upon consideration of the following detailed description, taken in conjunction with the accompanying drawings, in which like reference characters refer to like parts throughout, and in which:
For purposes of this disclosure, “full-duplex” communication should be considered to mean communication in which nodes transmit and receive at the same time, using at least some overlapping frequencies. Because there is usually some residual echo of the transmit signal in the receive path, each node has to perform echo cancellation. The received signal is:
r(t)=y(t)+a(t)+n(t)
where r(t) is the received signal, y(t) is the desired signal received from a distant transmitter, a(t) is the echo of the node's own transmit signal, and n(t) is noise. In receive mode, the node estimates a(t) and subtracts it from r(t). The echo signal is obtained via the convolution of the echo channel with the transmitted signal:
a(t)=x(t)*h(t)
where x(t) is the transmitted signal and h(t) is the echo channel. The echo channel may be estimated by the node by using a sounding signal and/or adaptive filtering techniques.
As noted above, many full-duplex systems operate in a point-to-point mode in which two nodes communicate between themselves. In such a case, Medium Access Control (MAC) is not required given that each of the two nodes can always transmit while simultaneously receiving from the other node. Examples of full-duplex point-to-point systems are Ethernet 1000BASE-T and 10GBASE-T systems.
Some full-duplex systems operate in a point-to-multipoint mode in which one node (sometimes referred to as a “master node” or “access point”) transmits to the other nodes and receives transmission from the other nodes. There is no direct communication between the other nodes. The master node or access point transmits in the downstream direction, and the other nodes share the upstream transmissions. The master node or access point is thus the only node that can initiate a full-duplex transmission.
In accordance with implementations of the subject matter of this disclosure, in a full-duplex system, each node in a plurality of nodes (which may include all nodes) is configured to initiate full-duplex communication with each other node. In effect, each node in the plurality of nodes can be a master node. A system in accordance with the subject matter of this disclosure typically includes up to 32 nodes. Such a system may be a home network, a small office network, or an access system.
As seen in
Although paths 101-106 are shown in
In order to avoid collisions, and to allow each node to estimate the echo channel for echo cancellation (i.e., as noted above, to cancel from a transmission in one direction echoes remaining from a transmission in the other direction), a control protocol, such as a Medium Access Control (MAC) protocol, for system 100 should control when and how a node may initiate full-duplex communication on a channel. In one implementation, a MAC protocol such as, e.g., a CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) protocol is used to access a channel. A node, which may be referred to as a “primary transmitter,” that is initiating a full-duplex communication broadcasts a signal to access the communications medium. This signal, which may be referred to as a Full-Duplex Start (FDS) signal, reserves the channel and identifies the other node that will form the full-duplex pair. All other nodes in system 100 also receive this signal.
Once the other node that is identified in the FDS (also referred to as the “secondary transmitter”) receives the FDS, the secondary transmitter knows that it is the other member of the full-duplex pair and can also transmit. From that point on, the other nodes in system 100, which also process the FDS (see below in connection with
In a variation, a similar scheme could be used to implement three-way full-duplex communication among three nodes, in which Node A transmits to node B, and node C transmits back to node A. This is a less common case than two-way full duplex, but it is possible in some scenarios. In such a scheme, the FDS identifies node C as the secondary transmitter, but the header of messages transmitted by the primary transmitter indicates that node B is the intended receiver of data transmitted by node A. In order for such an arrangement to work, node B must not receive the signal transmitted by node C. That is, node B and node C must be selected and configured so that they do not interfere each other. For example, in a frequency-division multiplexing environment, they could be configured to use different subcarriers.
A method 200 for initiating a full-duplex transmission between nodes in a network system, according to an implementation of the subject matter of this disclosure using an FDS signal, is diagrammed in
In one variant of a signaling format 300 used in accordance with the subject matter of this disclosure, shown in
In another variant of a signaling format 400 used in accordance with the subject matter of this disclosure, shown in
In a second implementation of the subject matter of this disclosure, based on time-division multiplexing, different nodes are assigned temporal slots during which they can engage in full-duplex communication. The assignments may be referred to as a “Medium Access Plan” (MAP), which may be promulgated by one of the nodes that is tasked with that function. In one variant of such an implementation, nodes are assigned to slots by pairs, and in each slot, the pairs assigned to that slot engage in full-duplex communication (unless neither node has information to transmit, in which case the slot remains unused). The assigning of temporal slots is made according to an analysis of which nodes are likely to need to transmit to which other nodes. Such an analysis may include review of previous resource reservation requests.
Normally, the MAP remains in effect until a new MAP is sent. Therefore, in order to ensure that every possible pair of nodes that may need to communicate can communicate, then unless the system includes a resource request mechanism, the MAP should include a slot for all possible pairings. However, in all but the smallest systems, assigning a temporal slot for every possible pair of nodes is impractical. One solution may be for the MAP to include a repeating pattern of slots that does not include every possible pairing, but on every nth repetition, where n is determined by the system design, the pattern is expanded to include all possible pairings.
In a second variant of a time-division multiplexing implementation, the MAP assigns temporal slots to individual nodes as primary transmitters. Upon the opening of any temporal slot, the primary transmitter assigned to that temporal slot begins full-duplex communication with a desired other node by sending an FDS identifying that other node. The full-duplex communication session lasts only until the temporal slot closes, unless both of the participating nodes require less than the full duration of the temporal slot to transmit whatever information they have to transmit. If the full-duplex communication session does not occupy the full duration of the temporal slot, the channel remains inactive for the balance of the temporal slot.
Returning to the first variant of a time-division multiplexing implementation in which each temporal slot is assigned to a particular pair of nodes, in one version of such a variant of a time-division multiplexing implementation, the two nodes assigned to a temporal slot begin transmitting immediately upon opening of that temporal slot. In a second version of such a variant of a time-division multiplexing implementation, upon opening of a temporal slot, even though the temporal slot is assigned exclusively to a particular pair of nodes, no transmission occurs until one of the nodes assigned to that temporal slot sends an FDS signal as in the first implementation, as discussed above. In this version of such a variant, if no FDS signal is sent by either node in a particular temporal slot, no transmission occurs during that temporal slot.
A method 500 for initiating a full-duplex transmission according to a time-division multiplexing implementation of the subject matter of this disclosure is diagrammed in
If at 503 it is determined that FDS signaling is in use, then at 506, one of the nodes assigned to the temporal slot, which has been designated “primary,” sends an FDS signal. At 507, both the primary node assigned to the temporal slot and the other node with which the primary nodes is communicating (the other node may or may not also be assigned to the temporal slot, depending on the particular variant that is implemented as discussed above) wait for a predetermined duration. At 508, both of the nodes that are communicating in the temporal slot node begin full-duplex transmission for a duration determined by signaling. At the conclusion of the temporal slot at 505, the two nodes stop transmitting (if the transmission has not concluded earlier) and flow returns to 502 for the next temporal slot.
In some variants of the first implementation based on FDS signaling, nodes not participating in the full-duplex transmission also need to know its duration, so they can contend for access to the channel when the full-duplex transmission is finished. Where transmission duration is not known ahead of time, because it is based on the duration of the longest of both transmissions, each of the two nodes that are participating in the full-duplex transmission has to signal the duration in an orthogonal way that can be understood by all nodes, including nodes that cannot separate the two transmitted signals. According to one mechanism, where the communication uses orthogonal frequency-division multiplexing (OFDM), each of the two nodes involved in a full-duplex transmission could transmit the duration information in a subcarrier that is not used by the other of the two nodes. If duration is determined solely by one of the nodes, then only that one node need signal the duration to other nodes. The use of such orthogonal transmissions to signal channel availability allows any half-duplex nodes (not shown) that may be present in the system to participate in the same MAC scheme along with full-duplex nodes.
Another situation that may be addressed by the use of OFDM signaling is a situation where echo cancellation cannot be performed, either because of low signal strength or where one node does not have much or any data to transmit to the other node. If OFDM is used, then rather than assigning all subcarriers in a channel to both nodes using the channel, the subcarriers in the channel can be divided up between the two nodes, so that there is no echo of transmissions by one node in the transmissions of the other node. Although such an arrangement would not be in-band full-duplex, it is still full-duplex in the sense defined above. Specifically, it is full duplex, in the sense of simultaneous bidirectional communication. And although different frequencies are used, in an OFDM context, and the gap between the subcarriers used in the different directions would not be large enough to allow use of external filters (e.g., diplexers). Therefore, there still should be precise time synchronization and coordination of transmissions between primary and secondary nodes, as provided using the FDS signaling described above.
G.hn OFDM communication implementations of the subject matter of this disclosure are described in concurrently-filed, commonly-assigned U.S. patent application Ser. No. ______, entitled “METHODS AND APPARATUS FOR PERFORMING FULL DUPLEX COMMUNICATIONS USING A G.hn PROTOCOL” (Attorney Docket No. MP10138/004048-0477-101), which is hereby incorporated by reference herein in its entirety.
As shown in
Thus it is seen that a MAC scheme, and a system implementing that scheme, in which any node can initiate full-duplex transmission with any other node, have been provided.
As used herein and in the claims which follow, the construction “one of A and B” shall mean “A or B.”
It is noted that the foregoing is only illustrative of the principles of the invention, and that the invention can be practiced by other than the described embodiments, which are presented for purposes of illustration and not of limitation, and the present invention is limited only by the claims which follow.
This claims the benefit of copending, commonly-assigned United States Provisional Patent Applications Nos. 62/382,908 and 62/382,913, both filed Sep. 2, 2016, each of which is hereby incorporated by reference herein in its respective entirety.
Number | Date | Country | |
---|---|---|---|
62382908 | Sep 2016 | US | |
62382913 | Sep 2016 | US |