Methods and apparatus for traffic management in a communication network

#The present invention relates to methods for management of traffic in a communication network comprising a first transport medium and a second transport medium connected in series with the first transport medium. The present invention also relates to a computer program product configured, when run on a computer, to carry out methods for management of traffic in a communication network, and to a network entity, Medium Access Controller and system for management of traffic in a communication network.

#Communication networks may employ direct or shared interconnections between entities or nodes in the network. In a direct interconnection, two entities of arbitrary type are directly coupled to one another via a full-duplex medium such as copper, which provides a full-time transmission path in each of the two possible directions in which signals can flow between the entities. As the interconnecting medium is dedicated to the entities, there is no sharing of the transport resource, or contention for its use. In a shared interconnection, two or more entities of arbitrary type are coupled to one another in a logical point-to-multipoint or point-to-point topology via a shared access medium. Examples of shared access media include IEEE 802.3 “Ethernet”, IEEE 802.11 “WiFi”, ITU G.984 GPON, LTE. Systems that employ arbitration of shared access media, including CSMA[/CD], GEM TC, DBA, etc., typically do so on the basis of requests and attempts to access a shared medium, as well as information about the medium itself.

#If multiple entities attempt to access a shared transport medium, contention builds, and entities are forced to wait for a “turn” to use the medium, or wait until one of their medium access attempts is successful and does not result in a “collision”. While the transmission subsystem of the entity is waiting for access to the medium, the subsystem generating the data to be transmitted may continue to produce data and pass it to the transmission subsystem. The transmission subsystem will typically attempt to buffer this data, meaning the data is stored in a temporary memory such as a FIFO, until it can be successfully emitted. Once this buffer memory is exhausted, the transmission subsystem will attempt to signal the generation subsystem to pause. Once buffer space is again available, or medium access is available, the transmission subsystem will signal the generation subsystem to resume.

#For a better understanding of the present invention, and to show more clearly how it may be carried into effect, reference will now be made, by way of example, to the following drawings in which:

#FIG. 1 illustrates a network scenario comprising two independent MAC domains;

#FIG. 2 is a flow chart illustrating process steps in a method for managing traffic in a communication network;

#In a network scenario according to examples of the present invention, in each domain of a heterogeneous transport network, at least one traffic client contends for capacity on a transport medium, and, in a majority of practical scenarios, a plurality of traffic clients contend for capacity on the transport medium. A centralised MAC controls access to the transport medium, receiving information from the at least one, or from at least some of the plurality, of traffic clients about their traffic demands, and granting transmission capacity on the transport medium to the various traffic clients. Each traffic client may manage one or more queues of data waiting to be transmitted on the relevant transport medium. A traffic client may have a dedicated buffer for each queue, the buffer comprising the contents of the queue, or a single buffer may contain all of the traffic in all of the queues of a single traffic client. An access grant made by a MAC for its transport medium authorises a traffic client of the MAC to transmit a certain amount of traffic on the transport medium. The access grant may specify that it only applies to traffic from the authorised client having a certain priority, or any one of a range of certain priorities. On receipt of an access grant, a traffic client transmits all or part of the contents of its buffers according to its own priorities or as allowed by the priorities specified in the access grant. For the sake of clarity, the following description refers to the making and implementing of individual grant decisions. However, it will be appreciated that in a majority of practical implementations, a plurality of grant decisions is assembled into a grant map, with individual grant decisions being fitted in to a computed or recomputed grant map. The individual grant decisions referred to in the above and following description may therefore be associated with other grant decisions in the construction of grant maps. Each grant decision forming a part of a grant map may be made and/or adjusted according to examples of the present invention.

#Examples of the present invention provide methods according to which a degree of coordination is achieved between traffic access to first and second transport media connected in series. This coordination is achieved via a reporting process in which a network entity receives not only traffic demand information from traffic clients of a second MAC, which controls access to a second transport medium, but also information about traffic demands from clients of a first MAC, which first MAC controls access to a first transport medium over which traffic will be forwarded to one of the traffic clients of the second MAC. The network entity then prepares a medium access grant decision for the second MAC, and may in some examples also prepare a guidance instruction for a medium access grant decision for the first MAC. With visibility not only of the traffic demand for clients of the second MAC, but also of traffic waiting at the first MAC to be transmitted to a traffic client of the second MAC, the network entity can ensure a coordination between access to the first and second traffic media, and so may be able to reduce at least one of a buffering time for traffic transmitting from the first to the second transport medium and/or an amount of traffic buffered between the first and second transport media. The network entity may in some cases be the second MAC or may be a separate orchestrating element, which may for example be realised in a control plane of an SDN architecture.

#It will be appreciated that the level of coordination achieved between traffic access to first and second transport media according to examples of the invention may vary, according for example to the nature of the information received by the network entity regarding traffic demands from clients the first MAC, and to whether and to what extent the network entity may influence or control medium access grants made by the first MAC. For example, if the network entity receives only limited information concerning traffic demands of clients of the first MAC, including a summary of poll results, or detailed poll results, without influence over the making of grant decisions for the first MAC, the network entity may not be in a position to know precisely how the first MAC will construct its map of grant decisions to process those demands. However, with the knowledge of what traffic is waiting to be transmitted via the first transport medium, the network entity can prepare for a worst case scenario. In this manner, the network entity may effectively implement a one way coordination at the second MAC, in which the network entity makes a medium access grant decision for the second MAC which accounts for a worst case scenario handling by the first MAC of its traffic demands. This one way coordination may for example involve a degree of over-provisioning, as is implemented in other aspects of network traffic management.