Method and apparatus for time-of-day synchronization between network nodes

Abstract

A node with local GPS time in an ad-hoc wireless network synchronizes with the network time immediately and slowly pulls the network time to GPS time. However, network nodes may immediately transition to the common GPS time after an island or group head node determines that the transition in network time does not disturb the communication links.

Description

GOVERNMENT LICENSE RIGHTS[0001]The U.S. Government may have a paid-up license in this invention and the right in limited circumstances to require the patent owner to license others on reasonable terms as provided for by the terms of contract DARPA SUO Program MDA972-99-9-0007.BACKGROUND OF THE INVENTION[0002]1. Technical Field[0003]The present invention pertains to synchronization of nodes within a network. In particular, the present invention pertains to a time-of-day synchronization (TOD) within wireless ad-hoc networks that determines an appropriate time for network nodes to switch to a common GPS time without disruption of existing communication links.[0004]2. Discussion of Related Art[0005]Commercial wireless devices employ clock synchronization techniques for their operation in WPAN (Wireless Personal Area Networks) and WLAN (Wireless Local Area Networks) type networks. Exemplary WPAN type devices include Bluetooth devices designed to support low bandwidth and short distance (...

AI Technical Summary

Benefits of technology

[0013]Accordingly, it is an object of the present invention to rapidly transition network nodes to GPS time within an ad-hoc wireless network.

[0018]When island member nodes receive the confirmation messages, the node local network time may be updated to GPS time in response to proper conditions. If acknowledgement messages are received from less than a predefined percentage of member nodes, the island head node retransmits a transition request message and repeats the negotiation for a predefined number of trials (e.g., typically three times). When the negotiation is not successful and the pulling interval is still large, the island head node delays by a random back-off interval and restarts the process again. In the case where some nodes do not transition to the GPS time, these nodes become Isolated and attempt to synchronize with the existing networks. The present invention technique may control degradation of the existing communication links based on the predefined value of the response percentage.

Problems solved by technology

The time synchronization problem is especially difficult in a tactical network due to the volatility of RF links.

This type of network includes extremely mobile, volatile, power and bandwidth limited operational conditions.

Due to the severe timing constraints imposed by the Small Unit Operation (SUO) security features, nodes must establish a common network time before commencing communications.

However, the time-of-day synchronized nodes constantly lose and regain connectivity with each other in a tactical or ad-hoc environment.

This potentially long pulling time by the flywheel technique may be unacceptable in the highly mobile, volatile, power and bandwidth limited tactical networks.

However, the sudden changes of network time may disturb the current communication in the existing network since nodes more than one hop away from a group or island head node may not timely receive information related to the GPS transition and be unaware of when to perform that transition to GPS time.

Further, if a node local clock time is not updated by neighboring nodes with GPS time, drifts in the local clock may accumulate during the long pulling period of the flywheel technique.

The accumulated clock drifts during the flywheel pulling interval may cause fragmented networks.

Unless time synchronization is achieved between these neighboring nodes, the merging of the fragmented networks cannot be resolved.

The time synchronization between fragmented networks is very expensive with respect to the hardware / software processing required.

Thus, pulling network time to GPS time by a node with GPS may be problematic in tactical ad-hoc networks due to a potentially long pulling time, accumulated clock drifts and very expensive hardware / software implementation for a merge.

Images