Unit 4

1
Mobile Transport Layer Mobile Computing
Unit-4
Unit-4: Mobile Transport Layer: Traditional TCP ,Indirect TCP ,Snooping TCP
,Mobile TCP, Fast retransmit/fast recovery ,Transmission /time-out freezing ,Selective
retransmission, Transaction oriented TCP
Traditional TCP
The Transmission Control Protocol (TCP) is one of the core protocols of the Internet protocol
suite, often simply referred to as TCP/IP. TCP is reliable, guarantees in-order delivery of data and
incorporates congestion control and flow control mechanisms.
TCP supports many of the Internet's most
popular application protocols and resulting
applications, including the World Wide Web,
e-mail, File Transfer Protocol and Secure
Shell. In the Internet protocol suite, TCP is the
intermediate layer between the Internet layer
and application layer.
The major responsibilities of TCP in an active
session are to:
• Provide reliable in-order transport of
data: to not allow losses of data.
• Control congestions in the networks: to not
allow degradation of the network performance,
• Control a packet flow between the
transmitter and the receiver: to not exceed
the receiver's capacity.
TCP uses a number of mechanisms to achieve high performance and avoid 'congestion collapse',
where network performance can fall by several orders of magnitude. These mechanisms control
the rate of data entering the network, keeping the data flow below a rate that would trigger
collapse. There are several mechanisms of TCP that influence the efficiency of TCP in a mobile
environment. Acknowledgments for data sent, or lack of acknowledgments, are used by senders
to implicitly interpret network conditions between the TCP sender and receiver.

2
Unit-4
Congestion Control
A transport layer protocol such as TCP has been designed for fixed networks with fixed
end- systems. Congestion may appear from time to time even in carefully designed networks. The
packet buffers of a router are filled and the router cannot forward the packets fast enough because
the sum of the input rates of packets destined for one output link is higher than the capacity of the
output link. The only thing a router can do in this situation is to drop packets. A dropped packet
is lost for the transmission, and the receiver notices a gap in the packet stream. Now the receiver
does not directly tell the sender which packet is missing, but continues to acknowledge all in-
sequence packets up to the missing one.
The sender notices the missing acknowledgement for the lost packet and assumes a packet
loss due to congestion. Retransmitting the missing packet and continuing at full sending rate
would now be unwise, as this might only increase the congestion. To mitigate congestion, TCP
slows down the transmission rate dramatically. All other TCP connections experiencing the same
congestion do exactly the same so the congestion is soon resolved.
Slow start
TCP’s reaction to a missing acknowledgement is quite drastic, but it is necessary to get rid of
congestion quickly. The behavior TCP shows after the detection of congestion is called slow
start. The sender always calculates a congestion window for a receiver. The start size of the
congestion window is one segment (TCP packet). The sender sends one packet and waits for
acknowledgement. If this acknowledgement arrives, the sender increases the congestion window
by one, now sending two packets (congestion window = 2). This scheme doubles the congestion
window every time the acknowledgements come back, which takes one round trip time (RTT).
This is called the exponential growth of the congestion window in the slow start mechanism.
But doubling the congestion window
is too dangerous. The exponential
growth stops at the congestion
threshold. As soon as the congestion
window reaches the congestion
threshold, further increase of the
transmission rate is only linear by
adding 1 to the congestion window
each time the acknowledgements
come back.

3
Unit-4
Linear increase continues
until a time-out at the sender
occurs due to a missing
acknowledgement, or until
the sender detects a gap in
transmitted data because of
continuous
acknowledgements for the
same packet. In either case
the sender sets the congestion
threshold to half of the
current congestion
window. The congestion window itself is set to one segment and the sender starts sending a
single segment. The exponential growth starts once more up to the new congestion threshold,
then the window grows in linear fashion.
Fast retransmit/fast recovery
The congestion threshold can be reduced because of two reasons. First one is if the sender
receives continuous acknowledgements for the same packet. It informs the sender that the
receiver has got all the packets upto the acknowledged packet in the sequence and also the
receiver is receiving something continuously from the sender. The gap in the packet stream is not
due to congestion, but a simple packet loss due to a transmission error. The sender can now
retransmit the missing packet(s) before the timer expires. This behavior is called fast retransmit.
It is an early enhancement for preventing slow-start to trigger on losses not caused by congestion.
The receipt of acknowledgements shows that there is no congestion to justify a slow start. The
sender can continue with the current congestion window. The sender performs a fast recovery
from the packet loss. This mechanism can improve the efficiency of TCP dramatically. The other
reason for activating slow start is a time-out due to a missing acknowledgement. TCP using fast
retransmit/fast recovery interprets this congestion in the network and activates the slow start
mechanism.
The advantage of this method is its simplicity. Minor changes in the MH’s software
results in performance increase. No changes are required in FA or CH.
The disadvantage of this scheme is insufficient isolation of packet losses. It mainly
focuses on problems regarding Handover. Also it effects the efficiency when a CH
transmits already delivered packets.

4
Unit-4
Problems with Traditional TCP in wireless environments
Slow Start mechanism in fixed networks decreases the efficiency of TCP if used with
mobile receivers or senders.
Error rates on wireless links are orders of magnitude higher compared to fixed fiber or
copper links. This makes compensation for packet loss by TCP quite difficult.
Mobility itself can cause packet loss. There are many situations where a soft handover
from one access point to another is not possible for a mobile end-system.
Standard TCP reacts with slow start if acknowledgements are missing, which does not
help in the case of transmission errors over wireless links and which does not really help
during handover. This behavior results in a severe performance degradation of an
unchanged TCP if used together with wireless links or mobile nodes
Classical TCP Improvements
Indirect TCP (I-TCP)
Indirect TCP segments a TCP connection into a fixed part and a wireless part. The following
figure shows an example with a mobile host connected via a wireless link and an access point to
the ‘wired’ internet where the correspondent host resides.
Standard TCP is used between the fixed computer and the access point. No computer in
the internet recognizes any changes to TCP. Instead of the mobile host, the access point now
terminates the standard TCP connection, acting as a proxy. This means that the access point is
now seen as the mobile host for the fixed host and as the fixed host for the mobile host. Between
the access point and the mobile host, a special TCP, adapted to wireless links, is used. However,
changing TCP for the wireless link is not a requirement. A suitable place for segmenting the
connection is at the foreign agent as it not only controls the mobility of the mobile host anyway
and can also hand over the connection to the next foreign agent when the mobile host moves on.

5
Unit-4
The foreign agent acts as a proxy and relays all data in both directions. If CH
(correspondent host) sends a packet to the MH, the FA acknowledges it and forwards it to the
MH. MH acknowledges on successful reception, but this is only used by the FA. If a packet is
lost on the wireless link, CH doesn’t observe it and FA tries to retransmit it locally to maintain
reliable data transport. If the MH sends a packet, the FA acknowledges it and forwards it to CH.
If the packet is lost on the wireless link, the mobile hosts notice this much faster due to the lower
round trip time and can directly retransmit the packet. Packet loss in the wired network is now
handled by the foreign agent.
Socket and state migration after handover of a mobile host
During handover, the buffered packets, as well as the system state (packet sequence number,
acknowledgements, ports, etc), must migrate to the new agent. No new connection may be
established for the mobile host, and the correspondent host must not see any changes in
connection state. Packet delivery in I-TCP is shown below:

6
Unit-4
Advantages of I-TCP

No changes in the fixed network necessary, no changes for the hosts (TCP protocol) necessary, all
current optimizations to TCP still work


Simple to control, mobile TCP is used only for one hop between, e.g., a foreign agent and mobile host

1. transmission errors on the wireless link do not propagate into the fixed network
2. therefore, a very fast retransmission of packets is possible, the short delay on the mobile
hop s known

It is always dangerous to introduce new mechanisms in a huge network without knowing exactly how
they behave.


New optimizations can be tested at the last hop, without jeopardizing the stability of the Internet.

It is easy to use different protocols for wired and wireless networks.

Disadvantages of I-TCP

Loss of end-to-end semantics:- an acknowledgement to a sender no longer means that a receiver
really has received a packet, foreign agents might crash.


Higher latency possible:- due to buffering of data within the foreign agent and forwarding to a new
foreign agent


Security issue:- The foreign agent must be a trusted entity

Snooping TCP
The main drawback of I-TCP is the segmentation of the single TCP connection into two TCP
connections, which loses the original end-to-end TCP semantic. A new enhancement, which
leaves the TCP connection intact and is completely transparent, is Snooping TCP. The main
function is to buffer data close to the mobile host to perform fast local retransmission in case of
packet loss.
Snooping TCP as a transparent TCP extension

7
Unit-4
Here, the foreign agent buffers all packets with destination mobile host and additionally
‘snoops’ the packet flow in both directions to recognize acknowledgements. The foreign agent
buffers every packet until it receives an acknowledgement from the mobile host. If the FA does
not receive an acknowledgement from the mobile host within a certain amount of time, either the
packet or the acknowledgement has been lost. Alternatively, the foreign agent could receive a
duplicate ACK which also shows the loss of a packet. Now, the FA retransmits the packet
directly from the buffer thus performing a faster retransmission compared to the CH. For
transparency, the FA does not acknowledge data to the CH, which would violate end-to-end
semantic in case of a FA failure. The foreign agent can filter the duplicate acknowledgements to
avoid unnecessary retransmissions of data from the correspondent host. If the foreign agent now
crashes, the time-out of the correspondent host still works and triggers a retransmission. The
foreign agent may discard duplicates of packets already retransmitted locally and acknowledged
by the mobile host. This avoids unnecessary traffic on the wireless link.
For data transfer from the mobile host with destination correspondent host, the FA
snoops into the packet stream to detect gaps in the sequence numbers of TCP. As soon as the
foreign agent detects a missing packet, it returns a negative acknowledgement (NACK) to the
mobile host. The mobile host can now retransmit the missing packet immediately. Reordering of
packets is done automatically at the correspondent host by TCP.
Snooping TCP: Packet delivery

8
Unit-4
Advantages of snooping TCP:

The end-to-end TCP semantic is preserved.


Most of the enhancements are done in the foreign agent itself which keeps correspondent host
unchanged.


Handover of state is not required as soon as the mobile host moves to another foreign agent.
Even though packets are present in the buffer, time out at the CH occurs and the packets are
transmitted to the new COA.


No problem arises if the new foreign agent uses the enhancement or not. If not, the approach
automatically falls back to the standard solution.

Disadvantages of snooping TCP

Snooping TCP does not isolate the behavior of the wireless link as well as I -TCP. Transmission
errors may propagate till CH.


Using negative acknowledgements between the foreign agent and the mobile host assumes
additional mechanisms on the mobile host. This approach is no longer transparent for
arbitrary mobile hosts.


Snooping and buffering data may be useless if certain encryption schemes are applied end-to-
end between the correspondent host and mobile host. If encryption is used above the transport
layer, (eg. SSL/TLS), snooping TCP can be used.

Mobile TCP
Both I-TCP and Snooping TCP does not help much, if a mobile host gets disconnected.
The M-TCP (mobile TCP) approach has the same goals as I-TCP and snooping TCP: to prevent
the sender window from shrinking if bit errors or disconnection but not congestion cause current
problems. M-TCP wants to improve overall throughput, to lower the delay, to maintain end-to-
end semantics of TCP, and to provide a more efficient handover. Additionally, M-TCP is
especially adapted to the problems arising from lengthy or frequent disconnections. M-TCP splits
the TCP connection into two parts as I-TCP does. An unmodified TCP is used on the standard
host-supervisory host (SH) connection, while an optimized TCP is used on the SH-MH
connection.
The SH monitors all packets sent to the MH and ACKs returned from the MH. If the SH
does not receive an ACK for some time, it assumes that the MH is disconnected. It then chokes
the sender by setting the sender’s window size to 0. Setting the window size to 0 forces the
sender to go into persistent mode, i.e., the state of the sender will not change no matter how long
the receiver is disconnected. This means that the sender will not try to retransmit data. As soon as
the SH (either the old SH or a new SH) detects connectivity again, it reopens the window of the
sender to the old value. The sender can continue sending at full speed. This mechanism does not
require changes to the sender’s TCP. The wireless side uses an adapted 8

9
Unit-4
TCP that can recover from packet loss much faster. This modified TCP does not use slow start,
thus, M-TCP needs a bandwidth manager to implement fair sharing over the wireless link.
Advantages of M-TCP:

It maintains the TCP end-to-end semantics. The SH does not send any ACK itself but forwards the
ACKs from the MH.


If the MH is disconnected, it avoids useless retransmissions, slow starts or breaking connections by
simply shrinking the sender’s window to 0.


As no buffering is done as in I-TCP, there is no need to forward buffers to a new SH. Lost packets
will be automatically retransmitted to the SH.

Disadvantages of M-TCP:

As the SH does not act as proxy as in I-TCP, packet loss on the wireless link due to bit errors
is propagated to the sender. M-TCP assumes low bit error rates, which is not always a valid
assumption.


A modified TCP on the wireless link not only requires modifications to the MH protocol
software but also new network elements like the bandwidth manager.

Transmission/time-out freezing
Often, MAC layer notices connection problems even before the connection is actually
interrupted from a TCP point of view and also knows the real reason for the interruption. The
MAC layer can inform the TCP layer of an upcoming loss of connection or that the current
interruption is not caused by congestion. TCP can now stop sending and ‘freezes’ the current
state of its congestion window and further timers. If the MAC layer notices the upcoming
interruption early enough, both the mobile and correspondent host can be informed. With a fast
interruption of the wireless link, additional mechanisms in the access point are needed to inform
the correspondent host of the reason for interruption. Otherwise, the correspondent host goes into
slow start assuming congestion and finally breaks the connection.
As soon as the MAC layer detects connectivity again, it signals TCP that it can resume
operation at exactly the same point where it had been forced to stop. For TCP time simply does
not advance, so no timers expire.
Advantages:

It offers a way to resume TCP connections even after long interruptions of the connection.


It can be used together with encrypted data as it is independent of other TCP mechanisms such as
sequence no or acknowledgements

Disadvantages:

Lots of changes have to be made in software of MH, CH and FA.

10
Unit-4
Selective retransmission
A very useful extension of TCP is the use of selective retransmission. TCP acknowledgements
are cumulative, i.e., they acknowledge in-order receipt of packets up to a certain packet. A single
acknowledgement confirms reception of all packets upto a certain packet. If a single packet is
lost, the sender has to retransmit everything starting from the lost packet (go-back-n
retransmission). This obviously wastes bandwidth, not just in the case of a mobile network, but
for any network.
Using selective retransmission, TCP can indirectly request a selective retransmission of
packets. The receiver can acknowledge single packets, not only trains of in-sequence packets.
The sender can now determine precisely which packet is needed and can retransmit it. The
advantage of this approach is obvious: a sender retransmits only the lost packets. This lowers
bandwidth requirements and is extremely helpful in slow wireless links. The disadvantage is that
a more complex software on the receiver side is needed. Also more buffer space is needed to
resequence data and to wait for gaps to be filled.
Transaction-oriented TCP
Assume an application running on the mobile host that sends a short request to a server from time
to time, which responds with a short message and it requires reliable TCP transport of the
packets. For it to use normal TCP, it is inefficient because of the overhead involved. Standard
TCP is made up of three phases: setup, data transfer and
release. First, TCP uses a three-way handshake to
establish the connection. At least one additional packet is
usually needed for transmission of the request, and
requires three more packets to close the connection via a
three-way handshake. So, for sending one data packet,
TCP may need seven packets altogether. This kind of
overhead is acceptable for long sessions in fixed networks,
but is quite inefficient for short messages or sessions in
wireless networks. This led to the development of
transaction-oriented TCP (T/TCP).
T/TCP can combine packets for connection establishment and
connection release with user data packets. This can reduce the number
of packets down to two instead of seven. The obvious advantage for
certain applications is the reduction in the overhead which standard
TCP has for connection setup and connection release. Disadvantage is
that it requires changes in the software in mobile host

11
Unit-4
and all correspondent hosts. This solution does not hide mobility anymore. Also, T/TCP exhibits
several security problems.
Classical Enhancements to TCP for mobility: A comparison

12
Unit-4
Assignment Questions
1. (a) Write brief notes on congestion control in traditional TCP.
(b) Compare several enhancements to TCP for mobility giving their relative advantages
and disadvantages
2. (a) Describe transaction oriented TCP.
(b) Explain Mobile TCP. How does a supervisory host send TCP packets to the mobile
node and to a fixed TCP connection.
3. (a) How does selective transmission improve the transmission efficiency? What are the
modifications required in the TCP receiver to implement the selective retransmission
protocol.
(b)Explain snooping TCP. What are it's advantages and disadvantages?
4. Describe indirect TCP. Explain the modifications of indirect TCP as the selective repeat
protocol and mobile - end transport protocol. What are the advantages and disadvantages
of indirect TCP?
5. (a) Why mobility results in packet loss?
(b) Compare the error rate in wired networks and mobile networks.
(c) Why we cannot change TCP completely just to support mobile users? What are the
consequences of it?
6. (a) What are the applications in which packet delayed is equivalent to packet lost?
Explain.
(b) What are the applications for which packet loss can create severe problems? Explain
it
7. (a) Why Access point maintains buffers in Indirect TCP?
(b) How Indirect TCP hides the problems of wireless links from fixed host?
(c) The foreign agent can act as a gateway to translate between the different protocols in
Indirect TCP. Comment

13
Mobile Computing
Unit-4
Database issues: Hoarding techniques, caching invalidation mechanisms.
client server computing with adapt ion , power-aware and context-aware
computing, transactional models, query processing, recovery, and quality of
service issues
A database is a collection of systematically stored records or information. Databases store data in
a particular logical manner. A mobile device is not always connected to the server or network;
neither does the device retrieve data from a server or a network for each computation. Rather, the
device caches some specific data, which may be required for future computations, during the
interval in which the device is connected to the server or network. Caching entails saving a copy
of select data or a part of a database from a connected system with a large database. The cached
data is hoarded in the mobile device database. Hoarding of the cached data in the database
ensures that even when the device is not connected to the network, the data required from the
database is available for computing.
Database Hoarding
Database hoarding may be done at the application tier itself. The following figure shows a simple
architecture in which a mobile device API directly retrieves the data from a database. It also
shows another simple architecture in which a mobile device API directly retrieves the data from
a database through a program, for ex: IBM DB2 Everyplace (DB2e).
(a)API at mobile device sending queries and retrieving data from local database
(Tier 1)
(b)API at mobile device retrieving data from database using DB2e (Tier 1)
Database Issues

14
Mobile Computing Database Issues
Unit-4
Both the two architectures belong to the class of one-tier database architecture because the
databases are specific to a mobile device, not meant to be distributed to multiple devices, not
synchronized with the new updates, are stored at the device itself. Some examples are
downloaded ringtones, music etc. IBM DB2 Everyplace (DB2e) is a relational database engine
which has been designed to reside at the device. It supports J2ME and most mobile device
operating systems. DB2e synchronizes with DB2 databases at the synchronization, application,
or enterprise server
The database architecture shown below is for two-tier or multi-tier databases. Here, the
databases reside at the remote servers and the copies of these databases are cached at the client
tiers. This is known as client-server computing architecture.
(a)Distributed data caches in mobile devices
(b)Similar architecture for a distributed cache memory in multiprocessor
systems
A cache is a list or database of items or records stored at the device. Databases are
hoarded at the application or enterprise tier, where the database server uses business logic and
connectivity for retrieving the data and then transmitting it to the device. The server provides and
updates local copies of the database at each mobile device connected to it. The computing API at
the mobile device (first tier) uses the cached local copy. At first tier (tier 1), the API uses the
cached data records using the computing architecture as explained above. From tier 2 or tier 3,
the server retrieves and transmits the data records to tier 1 using business logic and synchronizes
the local copies at the device. These local copies function as device caches.

15
Unit-4
The advantage of hoarding is that there is no access latency (delay in retrieving the
queried record from the server over wireless mobile networks). The client device API has
instantaneous data access to hoarded or cached data. After a device caches the data distributed by
the server, the data is hoarded at the device. The disadvantage of hoarding is that the consistency
of the cached data with the database at the server needs to be maintained.
Data Caching
Hoarded copies of the databases at the servers are distributed or transmitted to the mobile devices
from the enterprise servers or application databases. The copies cached at the devices are
equivalent to the cache memories at the processors in a multiprocessor system with a shared main
memory and copies of the main memory data stored at different locations.
Cache Access Protocols: A client device caches the pushed (disseminated) data records from a
server. Caching of the pushed data leads to a reduced access interval as compared to the pull (on-
demand) mode of data fetching. Caching of data records can be-based on pushed 'hot records'
(the most needed database records at the client device). Also, caching can be based on the ratio of
two parameters—access probability (at the device) and pushing rates (from the server) for each
record. This method is called cost-based data replacement or caching.
Pre-fetching: Pre-fetching is another alternative to caching of disseminated data. The process of
pre-fetching entails requesting for and pulling records that may be required later. The client
device can pre-fetch instead of caching from the pushed records keeping future needs in view.
Pre-fetching reduces server load. Further, the cost of cache-misses can thus be reduced. The term
'cost of cache-misses' refers to the time taken in accessing a record at the server in case that
record is not found in the device database when required by the device API.
Caching Invalidation Mechanisms
A cached record at the client device may be invalidated. This may be due to expiry or
modification of the record at the database server. Cache invalidation is a process by which a
cached data item or record becomes invalid and thus unusable because of modification, expiry,
or invalidation at another computing system or server. Cache invalidation mechanisms are used
to synchronize the data at other processors whenever the cache-data is written (modified) by a
processor in a multiprocessor system, cache invalidation mechanisms are also active in the case
of mobile devices having distributed copies from the server.
A cache consists of several records. Each record is called a cache-line, copies of which can be
stored at other devices or servers. The cache at the mobile devices or server databases at any

16
Unit-4
given time can be assigned one of four possible tags indicating its state—modified (after
rewriting), exclusive, shared, and invalidated (after expiry or when new data becomes available)
at any given instance. These four states are indicated by the letters M, E, S, and I, respectively
(MESI). The states indicated by the various tags are as follows:
a) The E tag indicates the exclusive state which means that the data record is for internal use
and cannot be used by any other device or server.
b) The S tag indicates the shared state which indicates that the data record can be used by
others.
c) The M tag indicates the modified state which means that the device cache
d) The I tag indicates the invalidated state which means that the server database no longer
has a copy of the record which was shared and used for computations earlier.
The following figure shows the four possible states of a data record i at any instant in the server
database and its copy at the cache of the mobile device j.
Four possible states (M, E, S, or /) of a data record /at any instance at the server
database and device j cache
Another important factor for cache maintenance in a mobile environment is cache consistency
(also called cache coherence). This requires a mechanism to ensure that a database record is
identical at the server as well as at the device caches and that only the valid cache records are
used for computations.

17
Unit-4
Cache invalidation mechanisms in mobile devices are triggered or initiated by the server. There
are four possible invalidation mechanisms – Stateless asynchronous, stateless synchronous,
stateful asynchronous and stateful synchronous.
Stateless Asynchronous: A stateless mechanism entails broadcasting of the invalidation of the
cache to all the clients of the server. The server does not keep track of the records stored at the
device caches. It just uniformly broadcasts invalidation reports to all clients irrespective of
whether the device cache holds that particular record or not. The term 'asynchronous' indicates
that the invalidation information for an item is sent as soon as its value changes. The server does
not keep the information of the present state (whether Emi, Mmi, Smi, or Imi) of a data-record in
cache for broadcasting later. The server advertises the invalidation information only. The client
can either request for a modified copy of the record or cache the relevant record when data is
pushed from the server. The server advertises as and when the corresponding data-record at the
server is invalidated and modified (deleted or replaced).
The advantage of the asynchronous approach is that there are no frequent, unnecessary transfers
of data reports, thus making the mechanism more bandwidth efficient. The disadvantages of this
approach are—(a) every client device gets an invalidation report, whether that client requires that
copy or not and (b) client devices presume that as long as there is no invalidation report, the copy
is valid for use in computations. Therefore, even when there is link failure, the devices may be
using the invalidated data and the server is unaware of state changes at the clients after it sends
the invalidation report.
Stateless Synchronous This is also a stateless mode, i.e., the server has no information
regarding the present state of data records at the device caches and broadcasts to all client
devices. However, unlike the asynchronous mechanism, here the server advertises invalidation
information at periodic intervals as well as whenever the corresponding data-record at server is
invalidated or modified. This method ensures synchronization because even if the in-between
period report is not detected by the device due to a link failure, the device expects the period-end
report of invalidation and if that is not received at the end of the period, then the device sends a
request for the same (deleted or replaced). In case the client device does not get the periodic
report due to link failure, it requests the server to send the report.
The advantage of the synchronous approach is that the client devices receive periodic
information regarding invalidity (and thus validity) of the data caches. The periodic invalidation
reports lead to greater reliability of cached data as update requests for invalid data can be sent to the
server by the device-client. This also helps the server and devices maintain cache consistency
through periodical exchanges. The disadvantages of this mode of cache invalidation are—(a)
unnecessary transfers of data invalidation reports take place, (b) every client device gets an
advertised invalidation report periodically, irrespective of whether that client has a copy of the
invalidated data or not, and (c) during the period between two invalidation reports, the client

18
Unit-4
devices assume that, as long as there is no invalidation report, the copy is valid for use in
computations. Therefore, when there are link failures, the devices use data which has been
invalidated in the in-between period and the server is unaware of state changes at the clients after
it sends the invalidation report.
Stateful Asynchronous The stateful asynchronous mechanism is also referred to as the AS
(asynchronous stateful) scheme. The term 'stateful' indicates that the cache invalidation reports are
sent only to the affected client devices and not broadcasted to all. The server stores the information
regarding the present state (a record I can have its state as Emi, Mmi, Smi, or Imi) of each data-record at
the client device caches. This state information is stored in the home location cache (HLC) at the
server. The HLC is maintained by an HA (home agent) software. This is similar to the HLR at the
MSC in a mobile network. The client device informs the HA of the state of each record to enable
storage of the same at the HLC. The server transmits the invalidation information as and when the
records are invalidated and it transmits only to the device-clients which are affected by the
invalidation of data. Based on the invalidation information, these device-clients then request the
server for new or modified data to replace the invalidated data. After the data records transmitted by
the server modify the client device cache, the device sends information about the new state to the
server so that the record of the cache-states at the server is also modified.
The advantage of the stateful asynchronous approach is that the server keeps track of the state
of cached data at the client device. This enables the server to synchronize with the state of records at
the device cache and keep the HLC updated. The stateful asynchronous mode is also advantageous in
that only the affected clients receive the invalidation reports and other devices are not flooded with
irrelevant reports. The disadvantage of the AS scheme is that the client devices presume that, as long
as there is no invalidation report, the copy is valid for use in computations. Therefore, when there is a
link failure, then the devices use invalidated data.
Stateful Synchronous: The server keeps the information of the present state (Emi, Mmi, Smi, or
Imi) of data-records at the client-caches. The server stores the cache record state at the home
location cache (HLC) using the home agent (HA). The server transmits the invalidation
information at periodic intervals to the clients and whenever the data-record relevant to the client
is invalidated or modified (deleted or replaced) at the server. This method ensures
synchronization because even if the in-between period report is not detected by the device due to
a link failure, the device expects the period-end report of invalidation and if it is not received at
the end of the period, then the device requests for the same.
The advantage of the stateful synchronous approach is that there are reports identifying
invalidity (and thus, indirectly, of validity) of data caches at periodic intervals and that the server also
periodically updates the client-cache states stored in the HLC. This enables to synchronize with the
client device when invalid data gets modified and becomes valid. Moreover, since the invalidation
report is sent periodically, if a device does not receive an invalidation report after

19
Unit-4
the specified period of time, it can request the server to send the report. Each client can thus be
periodically updated of any modifications at the server. When the invalidation report is not received
after the designated period and a link failure is found at the device, the device does not use the
invalidated data. Instead it requests the server for an invalidation update. The disadvantage of the
stateful synchronous approach is the high bandwidth requirement to enable periodic transmission of
invalidation reports to each device and updating requests from each client device.
Data Cache Maintenance in Mobile Environments
Assume that a device needs a data-record during an application. A request must be sent to the
server for the data record (this mechanism is called pulling). The time taken for the application
software to access a particular record is known as access latency. Caching and hoarding the
record at the device reduces access latency to zero. Therefore, data cache maintenance is
necessary in a mobile environment to overcome access latency.
Data cache inconsistency means that data records cached for applications are not invalidated at
the device when modified at the server but not modified at the device. Data cache consistency
can be maintained by the three methods given below:
I. Cache invalidation mechanism (server-initiated case): the server sends invalidation reports
on invalidation of records (asynchronous) or at regular intervals (synchronous).
II. Polling mechanism (client-initiated case): Polling means checking from the server, the state
of data record whether the record is in the valid, invalid, modified, or exclusive state. Each
cached record copy is polled whenever required by the application software during
computation. If the record is found to be modified or invalidated, then the device requests for
the modified data and replaces the earlier cached record copy.
III. Time-to-live mechanism (client-initiated case): Each cached record is assigned a TTL (time-
to-live). The TTL assignment is adaptive (adjustable) previous update intervals of that record.
After the end of the TTL, the cached record copy is polled. If it is modified, then the device
requests the server to replace the invalid cached record with the modified data. When TTL is
set to 0, the TTL mechanism is equivalent to the polling mechanism.
Web Cache Maintenance in Mobile Environments
The mobile devices or their servers can be connected to a web server (e.g., traffic information
server or train information server). Web cache at the device stores the web server data and
maintains it in a manner similar to the cache maintenance for server data described above. If an
application running at the device needs a data record from the web which is not at the web cache,
then there is access latency. Web cache maintenance is necessary in a mobile environment to
overcome access latency in downloading from websites due to disconnections. Web cache
consistency can be maintained by two methods. These are:

20
Unit-4
I. Time-to-live (TTL) mechanism (client-initiated case): The method is identical to the one
discussed for data cache maintenance.
II. Power-aware computing mechanism (client-initiated case): Each web cache maintained at
the device can also store the CRC (cyclic redundancy check) bits. Assume that there are
N cached bits and n CRC bits. N is much greater than n. Similarly at the server, n CRC
bits are stored. As long as there is consistency between the server and device records, the
CRC bits at both are identical. Whenever any of the records cached at the server is
modified, the corresponding CRC bits at the server are also modified. After the TTL
expires or on-demand for the web cache records by the client API, the cached record
CRC is polled and obtained from the website server. If the n CRC bits at server are found
to be modified and the change is found to be much higher than a given threshold (i.e., a
significant change), then the modified part of the website hypertext or database is
retrieved by the client device for use by the API. However, if the change is minor, then
the API uses the previous cache. Since N » n, the power dissipated in the web cache
maintenance method (in which invalidation reports and all invalidated record bits are
transmitted) is much greater than that in the present method (in which the device polls for
the significant change in the CRC bits at server and the records are transmitted only when
there is a significant change in the CRC bits).
Client-Server Computing
Client-server computing is a distributed computing architecture, in which there are two types of
nodes, i.e., the clients and the servers. A server is defined as a computing system, which responds
to requests from one or more clients. A client is defined as a computing system, which requests
the server for a resource or for executing a task. The client can either access the data records at
the server or it can cache these records at the client device. The data can be accessed either on
client request or through broadcasts or distribution from the server.
The client and the server can be on the same computing system or on different computing
systems. Client-server computing can have N-tier architecture (N= 1, 2 ...). When the client and
the server are on the same computing system then the number of tiers, N = 1. When the client
and the server are on different computing systems on the network, then N = 2. A command
interchange protocol (e.g., HTTP) is used for obtaining the client requests at the server or the
server responses at the client.
The following subsections describe client-server computing in 2, 3, or N-tier
architectures. Each tier connects to the other with a connecting, synchronizing, data, or command
interchange protocol.

21
Unit-4
Two-tier Client-Server Architecture
Multimedia file server in two-tier client-server computing architecture (local copies 1 to j of
image and voice hoarding at the mobile devices)
The following figure shows the application server at the second tier. The data records are
retrieved using business logic and a synchronization server in the application server synchronizes
with the local copies at the mobile devices. Synchronization means that when copies of records
at the server-end are modified, the copies cached at the client devices should also be accordingly
modified. The APIs are designed independent of hardware and software platforms as far as
possible as different devices may have different platforms.
Three-tier Client-Server Architecture
In a three-tier computing architecture, the application interface, the functional logic, and the database
are maintained at three different layers. The database is associated with the enterprise server tier (tier
3) and only local copies of the database exist at mobile devices. The database connects to the
enterprise server through a connecting protocol. The enterprise server connects the complete
databases on different platforms, for example, Oracle, XML, and IBM DB2.

22
Unit-4
(a) Local copies 1 to j of database hoarded at the mobile devices using an enterprise database
connection synchronization server, which synchronizes the required local copies for application
with the enterprise database server (b) Mobile device with J2ME or BREW platform, APIs an OS
and database having local copies
Data records at tier 3 are sent to tier 1 as shown in the figure through a synchronization-cum-
application server at tier 2. The synchronization-cum-application server has synchronization and
server programs, which retrieves data records from the enterprise tier (tier 3) using business
logic. There is an in-between server, called synchronization server, which sends and
synchronizes the copies at the multiple mobile devices. The figure shows that local copies 1 to j
of databases are hoarded at the mobile devices for the applications 1 to j.

23
Unit-4
N-tier Client-Server Architecture
When N is greater than 3, then the database is presented at the client through in-between layers.
For example, the following figure shows a four-tier architecture in which a client device connects
to a data-presentation server at tier 2.
4-tier architecture in which a client device connects to a data-presentation server
The presentation server then connects to the application server tier 3. The application server can
connect to the database using the connectivity protocol and to the multimedia server using Java or
XML API at tier 4. The total number of tiers can be counted by adding 2 to the number of in-between
servers between the database and the client device. The presentation, application, and enterprise
servers can establish connectivity using RPC, Java RMI, JNDI, or HOP. These servers may also use
HTTP or HTTPS in case the server at a tier j connects to tier j+1 using the Internet.
Client-Server Computing with Adaptation
The data formats of data transmitted from the synchronization server and those required for the
device database and device APIs are different in different cases, there are two adapters at a
mobile device—an adapter for standard data format for synchronization at the mobile device and
another adapter for the backend database copy, which is in a different data format for the API at
the mobile device. An adapter is software to get data in one format or data governed by one
protocol and convert it to another format or to data governed by another protocol.

24
Unit-4
Figure shows an API, database, and adapters at a mobile device and the adapters at the
synchronization, application, or enterprise servers. Here the adapters are an addition used
for interchange between standard data formats and data formats for the API.
Transaction Models
A transaction is the execution of interrelated instructions in a sequence for a specific operation
on a database. Database transaction models must maintain data integrity and must enforce a set
of rules called ACID rules. These rules are as follows:

Atomicity: All operations of a transaction must be complete. In case, a transaction cannot be
completed; it must be undone (rolled back). Operations in a transaction are assumed to be
one indivisible unit (atomic unit).

Consistency: A transaction must be such that it preserves the integrity constraints and
follows the declared consistency rules for a given database. Consistency means the data is
not in a contradictory state after the transaction.

Isolation: If two transactions are carried out simultaneously, there should not be any
interference between the two. Further, any intermediate results in a transaction should be
invisible to any other transaction.
 Durability: After a transaction is completed, it must persist and cannot be aborted or
discarded. For example, in a transaction entailing transfer of a balance from account A to
account B, once the transfer is completed and finished there should be no roll back.

25
Unit-4
Consider a base class library included in Microsoft.NET. It has a set of computer software
components called ADO.NET (ActiveX Data Objects in .NET). These can be used to access the
data and data services including for access and modifying the data stored in relational database
systems. The ADO.NET transaction model permits three transaction commands:
1. BeginTransaction: It is used to begin a transaction. Any operation after
BeginTransaction is assumed to be a part of the transaction till the
CommitTransaction command or the RollbackTransaction
command. An example of a command is as follows:
connectionA.open();
transA = connectionA.BeginTransaction();
Here connectionA and transA are two distinct objects.
2. Commit: It is used to commit the transaction operations that were carried
out after the BeginTransaction command and up to this command. An
example of this is
transA.Commit();
All statements between BeginTransaction and commit must execute
automatically.
3. Rollback: It is used to rollback the transaction in case an exception is
generated after the BeginTransaction command is executed.
A DBMS may provide for auto-commit mode. Auto-commit mode means the transaction
finished automatically even if an error occurs in between.
Query Processing
Query processing means making a correct as well as efficient execution strategy by query
decomposition and query-optimization. A relational-algebraic equation defines a set of operations
needed during query processing. Either of the two equivalent relational-algebraic equations given
below can be used.

26
Unit-4
This means first select a column Contacts.cTelNum in a row in Contacts in which
Contacts.cTelNum column equals a column DialledNumbers.dTelNum by
crosschecking and matching the records of a column in Contacts with all the rows of
DialledNumbers. Then in the second step select the row in which Contacts.
firstChar = “R” and the selected cTelNum exists. Then in the third step project
cName and CTelNum.
This means that in first series of step, crosscheck all rows of Contacts and
DialledNumbers and select, after AND operation, the rows in which
Contacts.firstchar = “R” and Contacts.cTelNum =
DialledNumbers.dTelNum. Then in the next step project cName and cTelNum form
the selected records.
Query processing architecture
Π represents the projection operation, σ the selection operation, and Λ, the AND operation. It is
clear that the second set of operations in query processing is less efficient than the first. Query
decomposition of the first set gives efficiency. Decomposition is done by (i) analysis, (ii)
conjunctive and disjunctive normalization, and (iii) semantic analysis.
Efficient processing of queries needs optimization of steps for query processing. Optimization
can be based on cost (number of micro-operations in processing) by evaluating the costs of sets
of equivalent expressions. Optimization can also be based on a heuristic approach consisting of

27
Unit-4
the following steps: perform the selection steps and projection steps as early as possible and
eliminate duplicate operations.
The query optimizer employs (a) query processing plan generator and (b) query processing cost
estimator to provide an efficient plan for query processing.
Data Recovery Process
Data is non-recoverable in case of media failure, intentional attack on the database and
transactions logging data, or physical media destruction. However, data recovery is possible in
other cases. Figure below shows recovery management architecture. It uses a recovery manager,
which ensures atomicity and durability. Atomicity ensures that an uncommitted but started
transaction aborts on failure and aborted transactions are logged in log file. Durability ensures
that a committed transaction is not affected by failure and is recovered. Stable state databases at
the start and at the end of transactions reside in secondary storage. Transaction commands are
sent to the recovery manager, which sends fetch commands to the database manager. The
database manager processes the queries during the transaction and uses a database buffer. The
recovery manager also sends the flush commands to transfer the committed transactions and
database buffer data to the secondary storage. The recovery manager detects the results of
operations. It recovers lost operations from the secondary storage. Recovery is by detecting the
data lost during the transaction.
Recovery Management Architecture
The recovery manager uses a log file, which logs actions in the following manner:
1. Each instruction for a transaction for update (insertion, deletion, replacement, and
addition) must be logged.
2. Database read instructions are not logged
3. Log files are stored at a different storage medium.
4. Log entries are flushed out after the final stable state database is stored.

28
Unit-4
Each logged entry contains the following fields.

transaction type (begin, commit, or rollback transaction)


transaction ID


operation-type


object on which the operation is performed


pre-operation and post-operation values of the object.
A procedure called the Aries algorithm is also used for recovering lost data. The basic steps of
the algorithm are:
I. Analyse from last checkpoint and identify all dirty records (written again after operation
restarted) in the buffer.
II. Redo all buffered operations logged in the update log to finish and make final page.
III. Undo all write operations and restore pre-transaction values.
The recovery models used in data recovery processes are as follows:
I. The full recovery model creates back up of the database and incremental backup of the
changes. All transactions are logged from the last backup taken for the database.
II. The bulk logged recovery model entails logging and taking backup of bulk data record
operations but not the full logging and backup. Size of bulk logging is kept to the
minimum required. This improves performance. We can recover the database to the point
of failure by restoring the database with the bulk transaction log file backup. This is
unlike the full recovery model in which all operations are logged.
III. The simple recovery model prepares full backups but the incremental changes are not
logged. We can recover the database to the most recent backup of the given database.
QoS Issues:
Quality of service (QoS) mechanism controls the performance, reliability and usability of a
telecommunications service. Mobile cellular service providers may offer mobile QoS to customers just as
the fixed line PSTN services providers and Internet service providers may offer QoS. QoS mechanisms are
always provided for circuit switched services, and are essential for non-elastic services, for example
streaming multimedia. It is also essential in networks dominated by such services, which is the case in
today's mobile communication networks, but not necessarily tomorrow.
Mobility adds complication to the QoS mechanisms, for several reasons:

29
 A phone call or other session may be interrupted after a handover, if the new base station is
overloaded. Unpredictable handovers make it impossible to give an absolute QoS guarantee during a
session initiation phase.
 The pricing structure is often based on per-minute or per-megabyte fee rather than flat rate, and may
be different for different content services.
 A crucial part of QoS in mobile communications is grade of service, involving outage probability
(the probability that the mobile station is outside the service coverage area, or affected by co-channel
interference, i.e. crosstalk) blocking probability (the probability that the required level of QoS cannot
be offered) and scheduling starvation. These performance measures are affected by mechanisms such
as mobility management, radio resource management, admission control, fair scheduling, channel-
dependent scheduling etc.
Types
 Factors affecting QoS
 Measurement of QoS
 Cellular GoS
 Cellular audio quality
Factors affecting QoS
Many factors affect the quality of service of a mobile network.[
It is correct to look at QoS mainly
from the customer's point of view, that is, QoS as judged by the user. There are standard metrics of QoS to
the user that can be measured to rate the QoS. These metrics are: the coverage, accessibility (includes GoS),
and the audio quality. In coverage the strength of the signal is measured using test equipment and this can
be used to estimate the size of the cell. Accessibility is about determining the ability of the network to
handle successful calls from mobile-to-fixed networks and from mobile-to-mobile networks. The audio
quality considers monitoring a successful call for a period of time for the clarity of the communication
channel. All these indicators are used by the telecommunications industry to rate the quality of service of a
network.
Measurement of QoS
The QoS in industry is also measured from the perspective of an expert (e.g. teletraffic engineer).
This involves assessing the network to see if it delivers the quality that the network planner has been
required to target. Certain tools and methods (protocol analysers, drive tests and Operation and Maintenance
measurements), are used for this QoS measurement:

30
 Protocol analysers are connected to BTSs, BSCs, and MSCs for a period of time to check for
problems in the cellular network. When a problem is discovered the staff can record it and it can be
analysed.
 Drive tests allow the mobile network to be tested through the use of a team of people who take the
role of users and take the QoS measures discussed above to rate the QoS of the network. This test
does not apply to the entire network, so it is always a statistical sample.
 In the Operation and Maintenance Centres (OMCs), counters are used in the system for various
events which provide the network operator with information on the state and quality of the network.
 Finally, customer complaints are a vital source of feedback on the QoS, and must not be ignored.
Cellular GoS
In general, grade of service (GoS) is measured by looking at traffic carried, traffic offered and
calculating the traffic blocked and lost. The proportion of lost calls is the measure of GoS. For cellular
circuit groups an acceptable GoS is 0.02. This means that two users of the circuit group out of a hundred
will encounter a call refusal during the busy hour at the end of the planning period. The grade of service
standard is thus the acceptable level of traffic that the network can lose. GoS is calculated from the Erlang-B
formula, as a function of the number of channels required for the offered traffic intensity.
Cellular audio quality
The audio quality of a cellular network depends on, among other factors, the modulation scheme
(e.g., FSK, QPSK) in use, matching to the channel characteristics and the processing of the received signal
at the receiver using DSPs.

Unit 4

More Related Content

What's hot(20)

Similar to Unit 4(20)

Recently uploaded(20)

Unit 4