An overview of ETX

Here is a overview of the expected transmission count metric (ETX).

The earlier metric most commonly used by existing ad hoc routing protocols is minimum hop-count. It assumes that link quality is a binary concept, either the link exists or not. While often true in wired networks, this is not a reasonable approximation in the wireless case: it should the quality of wireless links takes into account.

There are several facets of link quality:

• Link loss ratios: The minimum hop-count routes sometimes are slow because they include links with high loss ratios, which cause bandwidth to be consumed by retransmissions. A two-hop path over reliable or fast links can exhibit better performance than a one-hop path over a lossy or slow link.
• Asymmetric: Because 802.11b uses link-level ACKs to confirm delivery, both directions of a link must work well in order to avoid retransmissions. Since most nodes in the network are involved in at least one asymmetric link, routing protocols must cope with asymmetry to be effective.
• Channel contention
• Link load
• Queuing delay
• Link data rate

Minimum hop-count does not take these factors into account, furthermore it also ignore the interference between successive hops of multi-hop paths.

The ETX of a link is the predicted number of data transmissions required to send a packet over that link, including retransmissions. The ETX of a route is the sum of the ETX for each link in the route. Since each attempt to transmit a packet can be considered a Bernoulli trial, the ETX of a link is the reciprocal of d_f × d_r, which the d_f is delivery ratio and t d_r is the reverse delivery ratio. The delivery ratios d f and dr are measured using dedicated link probe packets.

ETX has several important characteristics:

• ETX is based on delivery ratios, which directly affect throughput.
• ETX detects and appropriately handles asymmetry by incorporating loss ratios in each direction.
• ETX can use precise link loss ratio measurements to make fine-grained decisions between routes.
• ETX penalizes routes with more hops, which have lower throughput due to interference between different hops of the same path .
• ETX tends to minimize spectrum use, which should maximize overall system capacity.
• Since each node broadcasts the probe packets instead of unicasting them, the probing overhead is substantially reduced.
• ETX suffers little from self-interference since we are not measuring delays.

ETX’s improvement over HOP is more pronounced at longer path lengths. And if several TCP transfers are carried out between the same pair of nodes at different times, they should yield similar throughput using ETX, while the throughput under HOP will be more variable.

ETX also have some disadvantages:

• Since broadcast probe packets are small, and are sent at the lowest possible data rate, they may not experience the same loss rate as data packets sent at higher rates.
• ETX does not directly account for link load or data rate. A heavily loaded link may have very low loss rate, and two links with different data rates may have the same loss rate.

Several aspects of ETX could be improved in the future: its predictions of loss ratios for different packet sizes, particularly for 802.11bACKs; its handling of networks with links that run at a variety of bit-rates; and the robustness of ETX probes when competing with high levels of data traffic.