Communication Delay. The communication delay between directly connected (adjacent) nodes is expressed in terms of the minimum event-response delay, D, and network imprecision, d. These parameter are measured at the network level. A message broadcast by a node at real time t is expected to arrive at its directly connected adjacent nodes, be processed, and subsequent messages to be generated by those nodes within the time interval [t+D, t+D+d]. Communication between independently- clocked nodes is inherently imprecise. The network imprecision, d, is the maximum time difference among all receivers of a message from a transmitting node with respect to real time. The imprecision is due to many factors including, but not limited to, the drift of the oscillators with respect to real time, jitter, discretization error, temperature effects and differences in the lengths of the physical communication media. These parameters are assumed to be bounded, D > 0, d ≥ 0, and both have units of real-time clock ticks and their values known in the network. The communication delay, denoted , is expressed in terms of D and d, is defined as = D+d, and has units of real-time clock ticks. In other words, we assume synchronous communication and bound the communication delay between any two directly connected adjacent nodes by [D, ]. However, for simplicity of notation, in the remainder of this paper we assume that the messages arrive logically at the same time at the destination nodes.
Communication Delay. Due to the reliable group communication platform, network delay is amplified by the necessary acknowledgments between the group members. The speed of light puts a lower bound on the minimum network delay. For example, a laser pulse that travels through a fiber optic cable takes ≈ 10 ms to travel from New York to San Francisco, ≈ 21 ms from Paris to San Francisco, and ≈ 40 ms from London to Sydney. In practice, networks today are about 3 to 4 times slower than the lower bound. To put this into perspective, an 850MHz Pentium III PC performs a single 512-bit modular exponentiation (one of the most expensive, but most basic public key primitives) in under 1 ms. Moreover, the speed of computers continue to increase. Comparing this with the WAN network delay, it is clear that reducing the number of communication rounds is much more important in the long run for an efficient group key agreement scheme than reducing the computation overhead.
Communication Delay. ≈ ≈ ≈ Due to the reliable group communication platform, network delay is amplified by the necessary acknowledgments between the group mem- bers. The speed of light puts a lower bound on the minimum network delay. For example, a laser pulse that travels through a fiber takes 10 ms between New York and San Francisco, 21 ms between Paris and San Francisco, and 40 ms from London to Sydney. In practice the networks today are slower than the lower bound by about a factor of 4 (due to switching overhead, etc.). To put this into perspective, an 850MHz Pentium III PC performs a single 512-bit modular exponentiation (one of the most expensive, but most basic public key primitives) in under 1 ms. Moreover, the speed of computers continue to increase. Comparing this with the WAN network delay, it is clear that reducing the number of communication rounds is much more important in the long run for an efficient group key agreement scheme than reducing the computation overhead.
Communication Delay. If Practice or a Provider does not respond to a Patient’s e-mail message within one (1) day, Patient agrees to use another means of communication to contact the Provider with respect to the issue. Neither Practice nor the Provider(s) will be liable to Patient for any loss, cost, injury or expense caused by, or resulting from, a delay in responding to Patient as a result of technical failures, including, but not limited to: (i) technical failures attributable to any internet service provider, (ii) power outages, failure of any electronic messaging software or failure to properly address e-mail mes- sages, (iii) failure of Practice’s or the Providers' computers or computer networks, or faulty telephone or cable data transmissions, (iv) any interception of e-mail communications by a third party; or (v) Patient’s failure to comply with the guidelines regarding use of e-mail communications set forth in this Agreement.
