Time Synchronization for Wireless Sensor Networks - Adaptive Signal Processing in Wireless Communications - page 387

• Clock correcting versus untethered clock

• Clock correcting : Where the clock function in memory is modified after

each run of the time synchronization process.
Untethered clock

• : Where every node maintains its own clock as it is, and
keeps a time-translation table relating its clock to other nodes' clocks.
Thus, instead of updating its clock constantly, each node translates the
time information in the data packets coming from other nodes to its own
clock by using the time-translation table.
Synchronization approach •

• Sender-receiver : Where one of two nodes, which are synchronizing with

each other, sends a time-stamp message while the other one receives it.
Receiver-receiver • : Where a reference node transmits synchronization sig-

nals and two synchronizing nodes receive these signals and record the
time of receptions (time stamps).
Receiver-only

• : Where a group of nodes can be simultaneously synchro-
nized by only listening to the message exchanges of a pair of nodes.
Pairwise synchronization versus network-wide synchronization •

• Pairwise synchronization : Where the protocols are primarily designed to

synchronize two nodes, although they usually can be extended to handle
synchronization of a group of nodes.
Network-wide synchronization

• : Where the protocols are primarily designed

to synchronize a large number of nodes in the network.

Additional classifications can be found in [4]. In the following, we will summarize the
existing time synchronization protocols based on the last category.

13.4.1 Pairwise Synchronization
13.4.1.1 Timing-Sync Protocol for Sensor Networks (TPSN)
TPSN [17] uses the two-way message exchange mechanism, as discussed in the sender-
receiver synchronization approach described in section 13.3.1, to achieve the synchroni-
zation between two nodes. With only one round of message exchanges, and without any
statistical model on the variable delay components X i (AP) and X i (PA) in (13.2) and (13.3), a
simple estimate for θ (AP) is proposed in [17] as

UV
i

−

ˆ θ (AP) = i

, (13. 30)

2

(P) . Notice that in the original form of TPSN, it
does not estimate clock skew; therefore, frequent application of TPSN is needed to keep
the clock offset between two nodes under a certain limit.

(P) - T 1, i (A) and V i  T 4, i (A) - T 3, i

where U i  T 2, i

13.4.1.2 Maximum Likelihood Estimation for Clock Offset Based
on Two-Way Message Exchanges
Assume the clock offset θ (AP) is constant for N rounds of message exchanges. If X i (AP) and
X i (PA) in (13.2) and (13.3) are exponentially distributed with the same unknown mean λ,