The increasing volume and importance of point-to-multipoint traffic in virtualized data centers means the deployment of IP multicast is increasingly attractive. However, concerns about the ability of switches and routers based on commodity hardware to support the conventional IP multicast control plane and data plane, especially when there are thousands of participants in the multicast group communication, results in the infrequent deployment of IP multicast in data centers. This paper discusses the evolution of data center architectures towards the virtualized architectures in which technologies such as VXLAN, VXLAN-GPE, GENEVE, STT and NVGRE are used to build emulated Layer 2 networks that will support multi-tenancy at scale. These technologies are described and compared in terms of a number of factors, with emphasis laid on the manner in which they support multicast. Lastly, innovative approaches that have been proposed to circumvent the obstacles to deploying multicast in the data center IP fabric are also discussed and evaluated.

GMPLS is viewed as an attractive intelligent control plane for different network technologies and graceful restart is a key technique in ensuring this control plane is resilient and able to recover adequately from faults. This paper studies the graceful restart mechanism recently proposed for a key GMPLS protocol, RSVP-TE. The performance of RSVP-TE graceful restart is evaluated in a variety of scenarios, with the impact of factors such as the number of data plane connections, the degree of the restarting node and the manner in which recovery messages are sent explored. The results obtained give significant insight into the behaviour and performance of RSVP-TE graceful restart and the findings are leveraged to propose techniques and guidelines to improve the performance of RSVP-TE graceful restart.

GMPLS is viewed as an attractive intelligent control plane for different network technologies and graceful restart is a key technique to ensure this control plane is resilient and able to recover adequately from faults. This paper analyses the graceful restart mechanism proposed for a key GMPLS protocol, RSVP-TE. A novel analytical model, which may be readily adapted to study other protocols, is developed. This model allows the efficacy of graceful restart to be evaluated in a number of scenarios. It is found that, unsurprisingly, increasing control message loss and increasing the number of data plane connections both increased the time to complete recovery. It was also discovered that a threshold exists beyond which a relatively small change in the control message loss probability causes a disproportionately large increase in the time to complete recovery. The interesting findings in this work suggest that the performance of graceful restart is worthy of further investigation, with emphasis being placed on exploring procedures to optimise the performance of graceful restart.

The proposed uses of the resource reservation protocol (RSVP) now extend beyond reserving resources in Internet protocol (IP) networks to being a generic signaling protocol for generalised multi-protocol label switching (GMPLS). In any implementation of RSVP, there are a number of discretionary timing parameters, the values of which affect the efficacy of RSVP in establishing and maintaining reservations/connections. This work frames the interactions between key RSVP timing parameters and performance metrics as a multi-objective optimisation problem which, due to its intractable nature, is tackled using a reputable multi-objective evolutionary algorithm. It is shown that this approach is a feasible means of exploring many of the innate tradeoffs in soft-state protocols such as RSVP. This approach facilitates an extensive comparison of a number of variants of RSVP: standard RSVP, RSVP featuring the recently standardised retransmission algorithm and two subsequent variants of this algorithm, supporting the asymmetric delivery of RSVP control messages. These RSVP variants are compared in terms of multiple performance metrics under a number of different exemplar network conditions, giving insight into their relative merits. Furthermore, the relative significance of the different timing parameters is investigated and their most expedient values determined.

The design of regular virtual topologies to facilitate optical packet switching in networks with arbitrary physical topologies is studied. Due to the intractable nature of the problem, two different artificial intelligence based heuristics are used to find favourable solutions expeditiously. The impact that the spread of the degree of physical topology nodes has on the quality of solutions obtained is investigated. Two important counter-intuitive findings emerge from this paper. Firstly, increasing the spread of the nodal degree in the physical topology leads to an improvement in the quality of initial non-optimised (or random) solutions. Secondly and conversely, increasing the spread of nodal degree has a detrimental impact on the final optimised solutions obtained. Both these surprising conclusions are explained by introducing novel estimates for the initial cost and the margin for improvement on this cost. The results produced in this paper are interesting and are applicable to the design of regular virtual topologies in generic physical topologies and additionally, suggest principles that are applicable to the wider area of telecommunication network design.

Optical packet switching over arbitrary physical topologies generally requires complex routing schemes coupled with contention resolving buffering. However, the relatively immature nature of optical logic devices and the limitations associated with optical buffering provide significant incentive to reduce the routing complexity and avoid optical domain contentions. This paper examines how the Manhattan Street Network (MSN) may be used to facilitate optical packet switching over arbitrary physical topologies. Several pertinent novel costs are introduced. Simulated annealing is used to deploy the MSN (near) optimally as a virtual topology in the physical topologies. Favourable and encouraging results are obtained, indicating that deploying the MSN is an attractive means to switch packets optically over arbitrary topologies.