Image Processing Reference
In-Depth Information
[KAT] found that for high offered loads in networks with mainly unacknowledged uplink traffic
the stable throughput saturation is at %. Nevertheless, with increasing load the success probability
decreases, so having the offered load between % and % is best.
Very important optimization parameters for IEEE .. are the lengths of beacon interval
and super frame. Low cycles mean low energy consumption but can result in high latencies and
low bandwidths [LKR]. Short frame sizes, in contrast reduce the available bandwidth by the
number of beacons, increased collisions at the beginning of a frame and packets deferred to the
next frame, when the remaining time is too short to it it in [KAT]. The highest throughput
when using the shortest frame sizes and  byte packets is  kbps for one device [LKR] or
 kbps in case of more than one device. [PRML] recommends to use low back-offs and short
super frame orders only, if the resulting short channel access time is critical and the offered load
is low. Otherwise the resulting large number of collisions is inefficient energy wise and leads to
poor utilization of the channel. That means, for each network the beacon interval and super frame
lengths need to be optimized individually, depending on the number of nodes and the desired
throughput.
Several other aspects though have also been examined for network optimization. [SMLW]
find that spatial diversity with the help of mesh topology helps to increase the performance, as
the latter is extremely location dependent. In [BGSV] the polling is optimized for respective
setups. In [HMBJ] the main focus is on the hidden node problem. It is found that hidden nodes
can significantly impact the performance within a network and that, depending on load, desired
throughput and beacon interval, an optimum transmit power exists. [KANS] introduce a queu-
ing system, in which for high-priority traffic (such as alarm reports, PAN management messages,
and GTS requests) a queue with different parameters (back-off, contention window length, etc.)
is used.
NoneofthepublicationsindictedthattheuseofIEEE../Zigbeefortheintendedusecases
is critical; just care needs to be taken when designing the network.
25.6 Coexistence of WPAN and WLAN
Multiple wireless technologies will in the near future coexist: Within an enterprise, e.g., WLAN tech-
nologies could be used for flexible access to large corporate data bases while WPAN technologies
handle specific tasks (and cellular systems the voice communication). Ideally, every (mobile) unit
would connect effortlessly using whatever technology is most suitable at the time. he smooth han-
dover between the technologies is thereby not the challenge one might expect. What causes problems
is, when the deployed communication technologies are placed in the same frequency band and/or
are linked by the application with each other (e.g., someone uses a Bluetooth headset with the mobile
telephone).
As mentioned, all three investigated technologies can be used in the . GHz ISM band. All can be
useful on the factory floor and might even be found in the same device or handle parts of the same
application.
Numerous publications cover the mutual interference and performance impairments of Bluetooth
and IEEE . (e.g., [Enn,Zyr,Fum,GvDS,How,Mob,MZ]). Depending on the inves-
tigated scenarios, the assessment of the situation varies from “good reliability even in fairly dense
environments” to “the effects of interference can be quite severe.” A relative agreement exists in the
parameters that determine the systems' performances: link distances (BT-BT, .b-.b, BT
transmitter-.b receiver, BT receiver-.b transmitter), traffic load, Bluetooth packet type,
density of units, local propagation conditions.
 
Search WWH ::




Custom Search