High availability survivable networks: When is reducing MTTR better than adding protection capacity?

In a survivable network, which withstands any single failure by design, service availability comes to depend primarily on dual-failure considerations. In this regard, most research has so far considered strategies for addition of further protection capacity, or post-failure reconfiguration of protection resources, to achieve full or partial dual-failure survivability. When considering outage due to dual-failure effects, however, reductions in the physical repair time of failures can also enhance availability. Thus, an interesting and possibly economic strategy may exist of trading off or balancing combined investments in MTTR reduction (to shorten outage times) and adding additional capacity (to increase the dual-failure restorability R2, to in effect prevent outages from occurring). Investments made by a network operator to improve service path availability could be directed to either approach, but a priori it is unclear which would be most effective, or in what framework an operator could approach this planning problem. Our contribution is to point out the nature of this trade-off and provide an analytical framework within which, if data is available, an operator could optimally allocate investment between MTTR reduction initiatives and capacity additions to enhance R2. Using plausible models for the behavior of the key functions we also demonstrate that such beneficial strategies exist and give further insight into the nature of the trade-off.