Full names of abbr?
CSMA: Carrier Sense Multiple Access
FDMA: Frequency-division multiple access
FHSS: Frequency hopping spread spectrum
WLAN: wireless local area network
WPAN: Wireless personal area network
SDMA: space devision multiple access
VLR: visitor location register
HLR: home location register
DSSS: direct sequence spread spectrum
LTE: Long-Term Evolution
AP: Access Point
LA: Local Area
DCF: Distributed Coordination Function
PCF: Point
ESS: Extended Service Set
Midterm
No. They are different.
Wireless applications can be run on mobile devices as well as from devices in fixed locations, as long as they make use of a wireless communication network.
Mobile applications can be run on a mobile device without accessing a wireless network. But in most cases, wireless applications are a subset of mobile applications, because even though mobile applications do not require wireless connectivity, they can often benefit from it when it is available.
Special characteristics of wireless communications, mobile devices and mobility:
Tut 03-3, Tut 03-4
\[Capacity = \frac{1}{2}\times \#cell \times \frac{channel\ capacity}{each\ channel\ bandwidth} \div Freq\ reuse\ factor\]
Analog modulation superimposes the baseband signal on the carrier signal so as to shift the data signal to a higher frequency. Purpose:
CSMA requires a node to first monitor the channel before it actually sends a packet. If the channel is free, the node will access the channel directly. If busy, the node will wait for the channel to become clear.
Hidden Terminal Problem happens when two nodes that are outside each other's transmission range are trying to send the packets to the same node at the same time. They cannot use carrier sense of carrier detection mechanism to avoid collision.
Exposed terminal problem happens when two nodes that are within each other's transmission range are trying to send the packets to different nodes that do not collide with each other. Two senders are within each other's range, one may detect anther's transmission and unnecessarily wait for accessing the medium.
RTS keeps all the nodes within sender's range quiet.
CTS keeps all the nodes within receiver's range quiet. → no HT.
If one only get RTS, no CTS, still transmit. → no ET.
Tut 03 2, Midterm
Carrier Sense
CS allows multiple mobile stations to access the wireless channel in a distributed way. Sense whether the channel is free before using.
Random back-off time
Collision Avoidance mechanism.
After failing, the station wait DIFS+randomized back-off time
Diff stations have diff back-off time
Dynamic contention window
The size of contention window determins the probability that two stations have the same back-off time.
The window can make a trade-off between the collision probability and unnecessary delay.
Residual back-off time
Fairness
Tries to access eailier have higher priority due to the shorter residual back-off time
ACK packet
Send from the receiver to sender
improve the wireless channel utilization
RTS/CTS packets
Reserve a wireless channel
Hidden terminal problem can be avoided
SIFS/PIFS/DIFS
Different priorities to different stations.
Priority: stations that intend to transmit
CTS/ACK/DATA > PCF > RTS
NAV time
Neighbours can know how long the channel is occupied. Will not try to access during the NAV time.
160921_24
What is an extended service set?
A set of Basic SS integrated together.
ESS appears the same to LLC layer as independent BSS
Mobile stations can move from one BSS to another transparently to LLC
What are these?
What is the difference between the DCF and PCF services?
DCF: transmit data packets based on a best effort strategy, asynchronous data service
PCF: based on a periodic polling strategy, time bounded service
Why do the standards specify different inter frame spaces?
Diff priority. CTS/ACK/DATA > PCF > RTS
For what purpose do the standards request the sender and receiver to exchange RTS and CTS packets before the actual data transmission?
Using RTS/CTS packets to reserve a wireless channel between the sender and receiver so that the hidden terminal problem can be avoided.
Three MAC methods used?
DCF CSMA/CA
DCF with RTS/CTS
PCF
Why do the standards set different NAV values in RTS and CTS packets?
RTS NAV time = s+s+s+CTS time + data time + ACK time
CTS NAV time = s+s+data time + ACK time
160921_24, Midterm
What is a piconet? What is master? What is slave?
Why does Bluetooth only support one master and up to seven active slaves in a piconet?
All active device in the piconet are assigned a 3-bit active member address 23 = 1+7
Why does Bluetooth use the frequency hopping technology?
To suppress interference
Why cannot a Bluetooth device become a master/master node in a scatternet?
If a device acts as masters in two Piconets, it would lead to identical behavior as both Piconets would have the same hopping sequence.
Why does the Bluetooth technology become more popular compared with infrared technology?
Bluetooth link does not have the line-of-sight restriction.
Infrared link requires two devices to directly see each other.
Which tech is used in Bluetooth for avoiding interferences? Why need?
FHSS, Frequency Hopping Spread Spectrum.
To avoid interferences at 2.4GHz ISM frequency spectrum.
Why need: 2.4GHz ISM frequency is free-of-charge, many applications, many WLANs and WPANs use at the same time. So interferences exist.
How does Bluetooth support the voice transfer and data transfer?
SCO(Synchronous connection oriented) for voice transfer.
ACL(Asynchronous connectionless) for data transfer.
170704_24, 04-9
Cannot assume continuous network connectivity and powerful client ability
Thin client → Resource limitation on mobile client forces moving client activities back to server side
Smart client → Client cannot be too thin since disconnection and low communication bandwidth of the network require some processing ability on client side for connection operations
Mobility of client causes additional problem for the server → when and where to transmit a reply back to client
How to solve? Extend C/S model to multiple tiers + conduct further optimizations
CAS
Pros
Client functions are shifted to agent → better for thin client
Only final result is transmitted to the client
Server can also shift some activities to agent
Agent can cache
Cons
Client app needs to change to communicate with agent
Do not support disconnection for client
CIS
Pros
Better for client with enough computational power and storage
No need to change client-side/service-side code
Diff protocols can be used between client ←> client agent; server ←> server agent
The pairs of agents masks disconnection and optimizes transmission
Cons
Client need more resources
Need to dev intercept agents
System overhead
P2P model is a generalized C/S model. No center server, no difference between client and server.
All nodes have the dual roles, serving as both S and C.
Nodes can communicate with each other directly
Nodes' operations are completely decentralized & asynchronous
Server can move and suffer from disconnection
170704_24, 04
Thin Client
Pro
Extend internet computing applications to mobile env
Need minimal to zero software deployment
Keep fresh data (connect to server all the time)
Provide high data security (all data in server)
Cons
Need persistent wireless connectivity
Performance is highly affected by network conditions
Availability (if server down, then down)
Application testing is difficult
Smart Client
Pros
Offline data access
Performance
Distributed computing → reduce server load
Security → have data both side
Cons
Application deployment ← diff mobile devices
Development complexity ← diff mobile devices with diff OS
Security risk ← mobile OS open to viruses
170704_24
Yes. Smart client application software is needed. To run the smart client application, the mobile devices should have following features:
Install a mobile OS
Connections over wireless networks
Persistent data storage
Computing ability
Can sync data with back-end data sources
Smartphones all have.
A conflict occurs when two or more users attempt to update the same data field with different values.
Detect
Both original and new values of data being updated must be stored.
When update, sync server check origin value from the client match the db one.
Not match, then conflicts
Resolve
Last-in wins
First-in wins
There are two main factors to affect the performance of location management. They are size of location area and location update methods.
For size of LA, big LA can decrease the location update cost and increase the paging cost. While smaller LA is different.
For the location update methods, static LAs may not match mobility changes and calling patterns. While dynamic LAs combines the use of location updates and paging operations, and make a tradeoff between them.
Three types of dynamic location update schemes: Time/Distance/Movement-based.
In routing optimization, the MN(mobile node) inform the CN(correspondent node) of its COA, then CN can send the packet directly to the MN.
Step 1: when the HA received packets for the MN that is not at home, the HA send binding update message to the CN. It has current COA of MN.
Step 2: CN caches COA and future packets are directly send to COA
Step 3: when MN moves to new FA, cached COA becomes stale
Then the CN sends a binding request message to the HA of the MN, HA send new binding update message to CN, CN cache again
Parallel paging
Reduce delay, good for paging with delay constraint
More bandwidth
Sequential paging
Save bandwidth, good for paging without bandwidth constraint
More delay
For TOA, the differences among the times must be very accurate. MS clock might shift so that error in distance calculation.
TDOA does not suffer from MS synchronization errors, MS sends out a signal, each station estimate the distance between the MS and the BS. Calculate difference between two distances.
TDOA only requires tight time synchronization among BSs → estimated distances are good approximations of the actual distances
160921_49
1 circle → at most n points
2 circles → at most 2 points
3 circles → only point