adaptive bandwidth management and joint call

8/4/2019 Adaptive Bandwidth Management and Joint Call

1/14


2/14

2 EURASIP Journal on Wireless Communications and Networking

algorithm deals with initial RAT selection only for new calls.Moreover, no analytical model is presented in the study.

Pillekeit et al. [8] propose a forced load balancing al-gorithm for heterogeneous UMTS/GSM network with colo-cated cells. In their approach, all the cells in the heteroge-neous network are classified into two groups: over-loaded

cells and under-loaded cells. The load balancing algorithmis triggered when a certain load threshold is exceeded in or-der to balance the traffic load in the heterogeneous network.However, the algorithm treats both new calls and handoffcalls alike. In practice, it is necessary to keep handoff calldropping probability below new call blocking probability.Moreover, no analytical model is presented in the study.

Romero et al. [9] propose a service-based RAT selectionpolicy for heterogeneous wireless networks. They illustratethe selection policy using heterogeneous network compris-ing GERAN and UTRAN, and a mix of voice and interactiveusers (e.g., www browsing). Examples of the service-basedselection policies are defined in the following [9].

(i) VG (voice GERAN) policy: this policy has only theservice class as input and allocates voice users into GERANand other services into UTRAN.

(ii) VU (voice UTRAN) policy: this policy acts in the op-posite direction as VG and allocates voice users to UTRANand interactive users to GERAN.

In the previous works mentioned above, no analyticalmodel has been developed for JCAC algorithms in orderto investigate connection-level QoS parameters in heteroge-neous cellular networks. Therefore, this paper models andanalyzes a JCAC algorithm in heterogeneous cellular net-works.

We propose adaptive bandwidth management and JCAC(AJCAC) scheme to enhance system utilization and connec-tion-level QoS in heterogeneous cellular networks support-ing multiple classes of calls such as voice and video. The pro-posed AJCAC scheme is designed to simultaneously achieve,the following objectives:

(1) distribute traffic load uniformly among available RATsto improve average system utilization,

(2) guarantee the QoS requirement of all admitted calls,(3) prioritize handoffcalls over new calls,(4) adapt the bandwidth of ongoing calls to improve

connection-level QoS and system utilization.

Uniform distribution of traffic load among multipleRATs in heterogeneous wireless network allows for a better

utilization of the radio resources. QoS requirements of alladmitted calls are guaranteed by allocating to each of thecalls at least the minimum bandwidth needed. Handoffcallsare prioritized over new calls by using different call rejectionthresholds for new and handoff calls, and also by using dif-ferent bandwidth adaptation mechanism for new and hand-offcalls. To the best of our knowledge, developing a schemethat achieves the above objectives at the same time in hetero-geneous wireless network is a novel work.

The contributions of this paper are twofold. Firstly,we combine adaptive bandwidth management and JCACscheme to enhance system utilization and connection-levelQoS in heterogeneous wireless networks. Secondly, we de-

RAT 1

RAT 2

1a

2aMT

1b

2b

1c

2cA group of

co-locatedcells

Figure 1: Two-RAT heterogeneous cellular networks with colocated

cells.

velop an analytical model for the AJCAC scheme, derive aver-age system utilization, new call blocking probability, handoffcall dropping probability, and examine the tradeoffbetweennew call blocking and handoffcall dropping.

The rest of this paper is organized as follows. Section 2presents the system model for heterogeneous wireless net-works. In Section 3, the components of the AJCAC schemeare described. Section 4 presents the Markov chain model ofthe AJCAC scheme. In Section 5, we investigate the perfor-mance of the AJCAC scheme through simulations.

2. SYSTEM MODEL AND ASSUMPTIONS

We consider a heterogeneous cellular network which con-sists of J number of RATs with colocated cells, similar to[7, 8]. Cellular networks such as GSM, GPRS, UMTS, andso forth can have the same and fully overlapped coverage,which is technically feasible, and may also save installationcost [10]. Figure 1 illustrates a two-RAT heterogeneous cel-lular network.

In heterogeneous cellular networks, radio resources canbe independently or jointly managed. We consider a situa-tion where radio resources are jointly managed in the het-erogeneous network and each cell in RAT-j (j = 1, . . . ,J)has a total of B j basic bandwidth units (bbu). The physicalmeaning of a unit of radio resources (such as time slots, codesequence, etc.) is dependent on the specific technological im-plementation of the radio interface [11]. However, no matterwhich multiple access technology (FDMA, TDMA, CDMA,or OFDM) is used, we could interpret system capacity interms of effective or equivalent bandwidth [1214]. There-fore, whenever we refer to the bandwidth of a call, we meanthe number of bbu that is adequate for guaranteeing the de-sired QoS for this call, which is similar to the approach usedfor homogeneous networks in [1416].

It is assumed that packet-level QoS is stochastically as-

sured by allocating at least the minimum effective band-width required to guarantee a given maximum probabilityon packet drop, delay, and jitter [17].

Our approach is based on decomposing heterogeneouscellular network into groups of colocated cells. As shown inFigure 1, cell 1a and cell 2a form a groupof colocated cells.Similarly, cell 1b and cell 2b form another group of colo-cated cells, and so on. Based on the following assumption-commonly made in homogeneous cellular networks, we as-sume that the types and amount of traffic are statistically thesame in all cells of each RATs [14, 15, 18, 19]. Therefore, thetypes and amount of traffic are statistically the same in allgroups of colocated cells.


3/14

O. E. Falowo and H. A. Chan 3

A newly arriving call will be admitted into one of the cellsin the group of colocated cells where the call is located. Whena mobile subscriber using a multimode terminal and havingan ongoing call is moving from one group of colocated cellsto another group of co-located cells, the ongoing call mustbe handed over to one of the cells in the new group of colo-

cated cells. For example (Figure 1), an ongoing call can behanded over from cell 2a to cell 2b or from cell 2a to cell 1b.Note that the handover consists of both horizontal and verti-cal handovers.

The correlation between the groups of colocated cellsresults from handoff connections between the cells of cor-responding groups. Under this formulation, each group ofco-located cells can be modeled and analyzed individually.Therefore, we focus our attention on a single group of colo-cated cells.

The heterogeneous network supports K classes of calls.Each class is characterized by bandwidth requirement, ar-rival distribution, and channel holding time. Each class-icall requires a discrete bandwidth value, b

i,w, where b

i,wbe-

longs to the set Bi = {bi,w} for i = 1,2, . . . , K and w =1,2, . . . , Wi. Wi is the number of different bandwidth valuesthat a class-i call can be allocated. bi,1 (also denoted as bi,min )and bi,Wi (also denoted as bi,max ) are, respectively, the min-imum and maximum bandwidth that can be allocated to aclass-i call. Note that bi,w < bi,(w+1) for i = 1,2, . . . , K andw = 1,2, . . . , (Wi 1).

The requested bandwidth of an incoming class-i call isdenoted by bi,req, where bi,req Bi. Let mi,j and ni,j denote,respectively, the number of ongoing new class-i calls andhandoff class-i calls, in RAT-j with 1 c mi,j (for newcalls) and 1 c ni,j (for handoff calls). Let bi,assigned c de-note the bandwidth assigned to call c of class-i in RAT-j inthe group of colocated cells, where bi,assigned c Bi. A call cof class-i is degraded ifbi,assigned c < bi,req whereas the call isupgraded ifbi,assigned c > bi,req.

If a class of calls(i.e., class-i calls) requires a fixed numberof bbu (i.e. constant bit-rate service), it becomes a specialcase in our model in which bi,min = bi,max and the set Bi hasonly one element. However, it will not be possible to upgradeor degrade this class of calls.

Following the general assumption in cellular networks,new and handoffclass-i calls arrive in the group of colocated

cells according to Poisson process with rate ni and hi , re-

spectively. The call holding time (CHT) of a class-i call is as-

sumed to follow an exponential distribution with mean 1/i[18, 19].

To characterize mobility, the cell residence time (CRT),that is, the amount of time during which a mobile terminalstays in a cell (same as the time, it stays in a group of colo-cated cells) during a single visit, is assumed to follow an ex-ponential distribution with mean 1/h, where the parameterh represents the call handoffrate. We assume that the CRT isindependent of the service class.

The channel holding time is the minimum of the CHTand the CRT. Because minimum of two exponentially dis-tributed random variables is also exponentially distributed[20], the channel holding time for new class-i calls, and

for handoff class-i call, is assumed to be exponentially dis-tributed with means 1/ni and 1/

hi , respectively.

Note that this set of assumptions has been widely usedfor homogeneous cellular networks in the literature, and isfound to be generally applicable in the environment wherethe number of mobile users is larger than the number of

channels [20].

3. PROPOSED AJCAC SCHEME

In this section, we describe the proposed AJCAC schemewhich consists of the following three components: joint calladmission controller, threshold-based bandwidth reserva-tion unit, and bandwidth adaptation (BA) controller. Thesecomponents are described in the following.

3.1. The joint call admission controller

The joint call admission controller implements the JCACalgorithm. The basic function of the JCAC algorithm is tomake call admission decision and uniformly distribute traf-fic load among all the available RATs in the network.

During call setup, a multimode mobile terminal request-ing a service sends a request to the joint call admission con-troller which implements the JCAC algorithm. The servicerequest contains the call type, service class, and bandwidthrequirements. The JCAC procedure is shown in Figure 2,where xi,j and yi,j denote, respectively, the residual bbu avail-able in RAT-j for new and handoff class-i calls. Whenevera call arrives, the JCAC attempts to allocate the maximum

bbu for this call (i.e., set bi,req=

bi,max ) provided that theavailable bbu in the selected RAT is greater than or equal tobi,max . If the available bbu in the selected RAT is less thanbi,max but greater than or equal to bi,req, the call will be as-signed a bandwidth between bi,req and bi,max . If the availablebbu is less than bi,req but greater than or equal to bi,1 (bi,min ),the call will be assigned a bandwidth between bi,1 and bi,req.If the available bbu in all the RATs is less than bi,1, BA al-gorithm (BAA) will be invoked to reduce the bandwidth ofsome ongoing call(s) in the chosen RAT. If the available bbuis still less than bi,1 for all available RATs, the call will be re-jected.

For new class-i calls, let Cni,j denote the total bbu available

in RAT-j, i,j the fraction of bbu available in RAT-j over thesummation of bbu available in all RATs, xi,j the residual bbuavailable in RAT-j, and Lni,j the current load in RAT-j. For

handoff class-i calls, the corresponding values are Chi,j , i,j ,

yi,j , and Lhi, j . Then

i,j =Cni,jJj=1

Cni,j

i, j,

Jj=1

i,j = 1 i.

(1)


4/14


Reject class-i call

Yes

n > JNo

n + +

No

bi,min yijYes

Admit class-i

call into the nth

RAT andallocate bi,min

Select the nth RATand apply BAA

n = 1

Yes

bi,req < bi,minNo

bi,req = bi,(req1)n = 1

Yes

n > JNo

n + +

No

bi,req yijYes

Admitclass-i

call in the

nth RATand allocate

bi,req

Select the nth RAT

Sort RATj for all j inincreasing order of current

load Lhij , and set n = 1

(Handoffcall) No

Arrival of class-i call

New call?Yes

Reject class-i call

Yes

n > JNo

n + +

No

bi,min xijYes

Admitclass-i

call into the

nth RATand allocate

bi,min

Select the nth RATand apply BAA

n = 1

Yes

bi,req < bi,minNo

bi,req = bi,(req1)n = 1

Yes

n > JNo

n + +

No

bi,req xijYes

Admitclass-i

call into the

nth RATand allocate

bi,req

Select the nth RAT

Sort RATj for all j inincreasing order of current

load Lnij , and set n = 1

Figure 2: Proposed adaptive load-based JCAC algorithm.

niJCAC

ni,1ni,2

ni,J

RAT 1

RAT 2

RAT J

Figure 3: Splitting of the arraival process in the group of colocatedcells.

Similarly,

i,j =Chi,jJj=1

Chi,j

i, j,

Jj=1

i,j = 1 i.

(2)

When a new or handoffcall arrives into a group of colocatedcells, the JCAC algorithm selects the least loaded RAT avail-able for the incoming call. The action of selecting a RAT foreach arriving new or handoff call in the group of colocated

cells leads to splitting of the arrival process. Figure 3 illus-trates the splitting of the arrival among J number of RATs inthe group of colocated cells.

As shown in Figure 3, the arrival rate in the group of colo-cated cells is split among all the available RATs. Each RAT hasa fraction of the arrival rate (ni ). Due to the uniform-load-distribution action of the JCAC algorithm, the mean arrivalrates of class-i calls into each RAT in the group of collocated

cells are as follows:

ni,j = i,jni i, j,

ni =J

j=1

ni,j i.(3)

Similarly

hi,j =i,jhi i, j,

hi =J

j=1

hi,j i,(4)


5/14


RAT 1 RAT 2 Accessnetworks

Available

bandwidth B1 = t21 t11

t01

B2 = t22t12

t02

Figure 4: Accessible bandwidth for a two-class, two-RAT system.

where ni and hi denote the arrival rates of new class-i calls

and handoff class-i calls, respectively, into the group of co-

located cells. ni,j and hi,j denote the arrival rates of new class-

i calls and handoffclass-i calls, respectively, into RAT-j in thegroup of colocated cells.

The arrival rates of a split Poisson process are also Pois-son [21]. Therefore, given that the mean arrival rate of class-i calls into the group of colocated cells is Poisson, the meanarrival rates of the split class-i calls into RAT-1, RAT-2, . . . ,

RAT-J are also Poisson.

3.2. Threshold-based bandwidth reservation unit

In order to maintain lower handoffdropping probability, thebandwidth reservation unit implements a bandwidth reser-vation policy that uses different thresholds for new and hand-offcalls. Figure 4 shows the bandwidth reservation policy fora two-class, two-RAT system.

The policy reserves bandwidth for aggregate handoffcalls, thus gives them priority over new calls. The policy alsoprioritizes among different classes of handoffcalls accordingto their QoS constraints by assigning a series of bandwidth

thresholds t1,j , t2,j , . . . , tk,j , for handoffcalls such that

t0,j t1,j ti,j t(i+1),j tk,j = Bj j,(5)

where t0,j denotes the total number of bbu available for allnew calls in RAT-j, and ti,j denotes the total number of bbuavailable for handoff class-i calls in RAT-j. Bj denotes thetotal number of bbu available in RAT-j.

3.3. Bandwidth adaptation controller

The bandwidth adaptation controller executes the BAAwhich is triggered when a new call arrives or when a call iscompleted. Most multimedia applications are adaptive. Forexample, voice can be encoded at 16 kbps, 32 kbps, 64kbps,and 128kbps by choosing appropriate encoding mecha-nisms. Similarly, video applications can be made rate adap-tive by using, for instance, a layered coding method. In layercoding method, the lowest layer (i.e., the base layer) con-tains the critical information for decoding the image se-quence at its minimum visual quality. Additional layers pro-vide increasing quality. All these encoded layersmay be trans-mitted when the network is underutilized. However, whenthe network resources are being fully utilized, only based

layer(s) which contain critical information may be transmit-ted.

As an illustration, if one would watch a 30-minutevideo clip encoded at 256 kbps and 64 kbps respectively. At256 kbps, one will see better pictures with better resolutionthan at 64 kbps. Therefore, the bandwidth adaptation af-

fects the quality of the real-time applications rather than thetransmission time. However, the minimum requested QoS ismaintained by ensuring that the bbu of the calls are not de-graded below the required minimum.

In the proposed AJCAC scheme, when the system is un-derutilized, all arriving new and handoffclass-i calls are ad-mitted by the JCAC algorithm with the highest bandwidthlevel (i.e., bi,max ) for the calls. This approach increases band-width utilization for the heterogeneous wireless network.However, when the network resources are being fully utilized,bandwidth adaptation controller is invoked to execute BAAon arrival of a new or handoffcall.

The BAA is triggered whenever there is a call arrivalevent or a call departure event. The BAA performs two mainprocedures: downgrades and upgrades ongoing calls. Thedowngrading procedure is activated in the arrival epoch (i.e.,when a new or handoff arrives to an overloaded group ofcolocated cells). BAA reduces the bandwidth of some on-going call(s) randomly selected in the system to free justenough bbu to accommodate the incoming call. Note thatan adaptive class-i call is never degraded below the mini-mum bbu necessary to guarantee its minimum QoS require-ments. The upgrading procedure is activated in the departureepoch.

In the arrival epoch, the BAA downgrading procedurecan be implemented in two ways. In the first implementa-tion, only ongoing new calls can be downgraded to accom-modate an incoming new call whereas both ongoing newand handoffcalls can be downgraded to accommodate an in-coming handoffcall. This approach further prioritizes hand-offcalls over new calls, in addition to the prioritization ob-tained by using different rejection thresholds for new andhandoff calls. In the second implementation, both new andhandoffcalls can be downgraded to accommodate an incom-ing new (or handoff) call. In this implementation, prioriti-zation of handoff calls over new calls can only be achievedby using different rejection threshold for new and handoffcalls.

In the departure epoch, when a call departs from a RATin the group of colocated cells, some of the ongoing call(s)

randomly selected in RAT of the group of colocated cells maybe upgraded by the BAA algorithm.

4. MARKOV CHAIN ANALYSIS OF THE AJCAC SCHEME

The AJCAC policy described in Section 3 is a multidimen-sional Markov chain. The state space of the group of co-located cells can be represented by a (2KJ)-dimensionalvector given as

= mi,j , ni,j : i = 1, . . . , k, j = 1, . . . ,J. (6)


6/14


The nonnegative integer mi,j denotes the number of on-going new class-i calls in RAT-j, and the nonnegative integerni,j denotes the number of ongoing handoff class-i calls inRAT-j. Let S denote the state space of all admissible states ofthe group of colocated cells as it evolves over time. An ad-missible state s is a combination of the numbers of users in

each class that can be supported simultaneously in the groupof colocated cells while maintaining adequate QoS and meet-ing resource constraints. The state S of all admissible states isgiven as

S =

=

mi,j , ni,j : i = 1, . . . , k, j = 1, . . . ,J

:

ki=1

mi,jc=1

bi,assignedc tn0,j j

ni,j

c=1

bi,assignedc thi,j i, j

ki=1

mi,jc=1

bi,assignedc +k

i=1

ni,jc=1

bi,assignedc Bj j

.

(7)

The constraints simply state that the sum of the band-width units of all admitted class-i calls cannot be more thanthe total bandwidth units available for that class of calls.Given that the system is in the current state, s, for the AJ-CAC scheme, the state transition could be triggered by anyof the following events.

(1) Admission of a new class-i call into RAT-j with thesuccessor state s+11 and transition rate q(s, s

+11 ). It follows that

q

s, s+11

=ni,j , s, s+11 S. (8)

(2) Admission of a handoff class-i call into RAT-j with thesuccessor state s+12 and transition rate q(s, s

+12 ). It follows that

q

s, s+12

=hi,j , s, s+12 S. (9)

(3) Departure of a new class-i call from RAT-j with thesuccessor state s11 and transition rate q(s, s11 ). It follows that

q

s, s11

= mi,j ni , s, s

11 S. (10)

(4) Departure of a handoff class-i call from RAT-j withthe successor state s12 and transition rate q(s, s

12 ). It follows

that

q(s, s12 ) = ni,j hi , s, s

12 S, (11)

where s, s+11 , s11 , s

+12 , and s

12 are the following matrices:

s =

m11 m1j m1J...

......

mi1 mij miJ...

..

.

..

.mK1 mK j mKJn11 n1j n1J

......

...ni1 ni j niJ

......

...nK1 nK j nKJ

,

s11 =

m11 m1j m1J...

......

mi1 mi j 1 miJ...

......

mK1 mK j mKJn11 n1j n1J

......

...ni1 ni j niJ

......

...nK1 nK j nKJ

,

s12 =

m11 m1j m1J...

......

mi1 mi j miJ...

......

mK1 mK j mKJn11 n1j n1J

......

...ni1 ni j 1 niJ

......

...nK1 nK j nKJ

.

(12)

The decision epochs are the arrival or departure of a newor handoff call. Joint call admission decisions are taken inthe arrival epoch. Every time a new or handoff class-i callarrives in the group of colocated cells, the JCAC algorithmdecides whether or not to admit the call, and in which RAT

to admit it. Note that call admission decision is made only atthe arrival of a call, and no call admission decision is madein the group of colocated cells when a call departs. When thesystem is in state s, an accept in RAT-j/reject decision mustbe made for each type of possible arrival, that is, an arrival ofa new class-i call, or the arrival of a handoffclass-i call in thegroup of colocated cells. The following are the possible JCACdecisions in the arrival epoch.

(1) Reject the class-i call (new or handoff) in the groupof collocated cells, in which case the state s does not evolve.

(2) Admit the class-i call into RAT-j without adaptingthe bandwidth of ongoing call(s) in the RAT, in which casethe state s evolves.


7/14


(3) Admit the class-i call into RAT-j after adapting thebandwidth of ongoing call(s) in the RAT, in which case states evolves.

Thus, the call admission action space A can be expressedas follows:

A = a = an1 , . . . , ank , ah1, . . . , ahk :ani , a

hi

0, 1, . . . , j,

j + 1

, . . . , J

,

i = 1, . . . , k

,

(13)

where ani denotes the action taken on arrival of a new class-icall within the group of colocated cells, and ahi denotes theaction taken on arrival of a handoffclass-i call from an adja-cent group of colocated cells. ani (or a

hi ) = 0 means reject the

new class-i (or handoffclass-i) call. ani (or ahi ) =+1 means ac-

cept the new class-i (or handoffclass-i) call into RAT-1 with-out adapting the bandwidth of existing call(s). ani (or a

hi ) =

1 means accept the new class-i (or handoffclass-i) call intoRAT-1 after adapting (degrading) the bandwidth of existing

call(s). a

n

i (or a

h

i )=

+j means accept the new class-i (or hand-off class-i) call into RAT-j without adapting the bandwidth

of existing call. ani (or ahi ) = j means accept the new class-i

(or handoff class-i) call into RAT-j after adapting (degrad-ing) the bandwidth of existing call(s).

In the departure epoch, the bandwidth adaptation unitmakes the decision to adapt (upgrade) or not to adapt thebandwidth of ongoing call(s). Thus, the call departure actionspace W can be expressed as follows:

W =

w =

0, 1

, (14)

where w = 0 means do not adapt the bandwidth of the ongo-ing call(s) and w = 1 means adapt the bandwidth of ongoing

call(s).Based on its Markovian property, the proposed AJCAC

scheme can be model as a (2KJ)-dimensional Markovchain. Let newi,j and hani,j denote the load generated by new

class-i calls and handoff class-i calls, respectively, in RAT-j.Then,

newi,j =ni,j

nii, j,

hani,j =hi,j

hii, j.

(15)

From the steady-state solution of the Markov model, per-formance measures of interest can be determined by sum-ming up appropriate state probabilities. Let P(s) denotes thesteady-state probability that the system is in state s (s S).From the detailed balance equation, P(s) is obtained as

P(s) =1

G

ki=1

Jj=1

newi,j

mi,jmi,j !

hani,j

ni,jni,j !

s S, (16)

where G is a normalization constant given by

G =sS

ki=1

Jj=1

newi,j

mi,jmi,j !

hani,j

ni,jni,j !

. (17)

4.1. New call blocking probability

A new class-i call is blocked in the group of colocated cells ifnone of the available RATs has enough bbu to accommodatethe new call with the minimum bandwidth requirement afterdegrading the ongoing new calls. Let Sbi S denote the setof states in which a new class-i call is blocked in the group ofcolocated cells. It follows that

Sbi =

s S :

bi,min +

ki

mi,jbi,min > tn0,j bi,min

+k

i=1

mi,j bi,min +k

i=1

ni,jc=1

bi,assignedc > Bj

j

.

(18)

Thus the new call blocking probability (NCBP), Pbi , for aclass-i call in the group of colocated cells is given by

Pbi = sSbi

P(s). (19)

4.2. Handoff call dropping probability

A handoff class-i call is dropped in the group of colocatedcells if none of the available RATs has enough bbu to accom-modate the handoff call with the minimum bandwidth re-quirement after degrading the ongoing new calls and handoffcalls. Let Sdi S denotes the set of states in which a handoffclass-i call is dropped in the group of colocated cells. It fol-lows that

Sdi =

s S :

1 + ni,j

bi,min > t

hi,j bi,min

+k

i=1

mi,j + ni,j

bi,min > Bj

j

.

(20)

Thus the handoff call dropping probability (HCDP) for aclass-i call, Pdi , in the group of colocated cells is given by

Pdi =

sSdi

P(s). (21)

4.3. Average system utilization

The average utilization of the heterogeneous wireless net-

work can be obtained by summing up for all the admissiblestate s (s S), the product of the system utilization in a par-ticular state s (s S), and the probabilityP(s) of the systembeing in that state. The average utilization Uof the heteroge-neous cellular network can be derived as follows:

U =sS

P(s)J

j=1

ki=1

mi,j

c=1

bi,assignedc +

ni,jc=1

bi,assignedc

. (22)

5. SIMULATION RESULTS

In this section, we evaluate the performance of the proposedAJCAC scheme with respect to new call blocking probability,


8/14


Table 1: Simulation parameters.

Parameters Class-1 call Class-2 call

Class-i call bbu set {2,3,4} {3,5,7}

Requested bbu (bi,req) 3 5

ni [1,8] [1,8]

i 0.5 0.5Other parameters

B1 B2 t0,1 t0,2 t1,1 t1,2 t2,1 t2,2 h

30 60 15 30 30 60 30 60 0.5

8765432

Call arrival rate

0

0.1

0.2

0.3

0.4

0.5

Newcal

lblockingprobability

Pb1 of AJCAC

Pb2 of AJCAC

Pb1 of NAJCAC

Pb2 of NAJCAC

Figure 5: Effectof varying the call arrival rate on the new call block-ing probability.

handoffcall dropping probability, and average system utiliza-tion. The results of the proposed AJCAC scheme are com-pared with that of the NAJCAC scheme. The system param-eters used are shown in Table 1.

The arrival rate of handoff class-i calls in the group ofcolocated cells is assumed to be proportional to the arrival

rate of new class-i calls byhi = (h/i)ni where h is the hand-

offrate.For comparison, we also model a JCAC algorithm with-

out adaptive bandwidth allocation in heterogeneous cellularnetwork and derive NCBP and HCDP for the nonadaptiveJCAC scheme.

Figures 5 and 6 show the performance of the AJCAC

scheme compared with that of NAJCAC. As shown inFigure 5, the NCBP of each class of calls increases with thecall arrival rate. The NCBP, Pb1 is always less than the NCBP,Pb2 because class-2 calls require more bbu than class-1 calls.Thus class-2 calls may be blocked due to insufficient bbu toaccommodate it whereas class-1 calls may still be acceptedinto the network. However, for both classes of calls, theNCBP for the AJCAC scheme is always less than the corre-sponding NCBP for the NAJCAC scheme. Note that lowerNCBP of the AJCAC scheme implies that its connection-level QoS is better than that of the NAJCAC scheme. Thereason why the NCBP of the AJCAC scheme is less thanthe NAJCAC scheme is as follows. When the total bbu al-

8765432

Call arrival rate

0

0.05

0.1

0.15

0.2

0.25

Handoffcalldroppingprobability

Pd1 of AJCAC

Pd2 of AJCAC

Pd1 of NAJCAC

Pd2 of NAJCAC

Figure 6: Effect of varying the call arrival rate on the handoff calldropping probability.

located to new calls is being fully utilized, incoming newcalls are rejected by the NAJCAC scheme whereas the AJ-CAC scheme adapts (degrades) the bandwidth of some ofthe ongoing adaptive calls to free just enough bbu to accom-modate the incoming new calls. Consequently, the NCBPof the AJCAC is less than that of the NAJCAC. However,an adaptive class-i call is never degraded below the mini-mum bbu necessary to guarantee its minimum QoS require-ments.

Figure 6 shows a similar trend for the HCDP for eachclass of calls, which increases with the call arrival rate. TheHCDP, Pd1 is always less than that the HCDP, Pd2, becauseclass-2 calls require more bbu than class-1 calls. However,for both classes of calls, the HCDP for the AJCAC schemeis always less than the corresponding HCDP for the NAJCACscheme. The reason why the HCDP of the AJCAC scheme isless than the NAJCAC scheme is as follows. When the systemis being fully utilized, incoming handoffcalls are rejected bythe NAJCAC scheme whereas the AJCAC scheme adapts (de-grades) the bandwidth of some of the ongoing adaptive callsto free just enough bbu to accommodate the incoming hand-offcalls. Consequently, the HCDP of the AJCAC is less thanthat of the NAJCAC.

Figures 7 and 8 compare NCBP and HCDP of the AJCACfor class-1 and class-2 call, respectively. One of the objectives

of the AJCAC scheme is to prioritized handoffcalls over newcalls. Figure 7 shows that the HCDP, Pd1 of the AJCAC isalways less than the Pb1. Similarly, it can be seen in Figure 8that the HCDP, Pd2 is always less that the NCBP, Pb2. Thisshows that handoff calls are prioritized over new calls. Thisprioritization of the handoff calls over new calls is achievedby making the handoff call rejection thresholds higher thanthe new call rejection thresholds.

Figures 9 and 10 show the effect of varying the new callrejection threshold, T0 on the NCBP and HCDP of the AJ-CAC and NAJCAC schemes for class-1 calls and class-2 calls,respectively. The additional system parameters used are asfollows: T01 = T0, T02 = 2T0, T0 = [0,30],

n1 =

n2 = 8. As


9/14


876543210

Call arrival rate

0

0.01

0.02

0.03

0.04

0.05

0.06

Callblocking/

droppingprobability

Pb1 of AJCACPd1 of AJCAC

Figure 7: Effect of varying the call arrival rate on the new callblocking probability and handoffcall dropping probability of class-1 calls.

876543210

Call arrival rate

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

Callblocking/droppingprobability

Pb2 of AJCAC

Pd2 of AJCAC

Figure 8: Effect of varying the call arrival rate on the new callblocking probability and handoffcall dropping probability of class-2 calls.

shown in Figure 9, at low threshold values, the NCPB, Pb1for the two JCAC schemes is high whereas the HCDP, Pd1 islow. As the threshold value, T0 increases, Pb1 decreases be-

cause new calls are given more access to the available band-width. On the other hand, the handoffdropping probability,Pd1 increases as a result of the higher degree of sharing be-tween the new and the handoffcalls. However, Pb1 and Pd1of the AJCAC are always less than the corresponding Pb1 andPd1 of the NAJCAC.

Figure 10 shows a similar trend for class-2 calls. At lowthreshold values, the NCPB, Pb2 for the two JCAC schemesis high whereas the HCDP, Pd2 is low. As the threshold value,T0 increases, Pb2 decreases whereas handoffdropping prob-ability, Pd2 increases. However, Pb2 and Pd2 of the AJCACare always less than the corresponding Pb2 and Pd2 of theNAJCAC.

302520151050

New call rejection threshold, T0

1E 04

1E 03

1E 02

1E 01

1E + 00

Callblocking/dr

oppingprobability

Pb1 of NAJCAC

Pd1 of NAJCAC

Pb1 of AJCAC

Pd1 of AJCAC

Figure 9: Effect of varying the new call rejection threshold, T0 onthe new call blocking probability and handoffcall dropping proba-bility of class-1 calls.

302520151050

New call rejection threshold, T0

1E 04

1E 03

1E 02

1E 01

1E + 00

Callblocking/droppingprobability

Pb2 of NAJCAC

Pd2 of NAJCAC

Pb2 of AJCAC

Pd2 of AJCAC

Figure 10: Effect of varying the new call rejection threshold, T0 onthe new call blocking probability and handoffcall dropping proba-bility of class-2 calls.

Figure 11 shows the normalized average system utiliza-tion for heterogeneous wireless network. The normalized av-erage system utilization of the AJCAC is higher that the nor-

malized average system utilization for the NAJCAC. The rea-son for improvement in system utilization of the AJCACscheme over NAJCAC scheme is as follows. When the sys-tem load is low, the AJCAC allocates maximum bbu to alladmitted calls, thereby improves the average system utiliza-tion whereas the NAJCAC allocates just the requested bbuto all admitted calls in the same class regardless of whetherthe traffic load is low or high. However, when the system isoperating at the full capacity, the AJCAC algorithm degradesthe bbu of some ongoing calls and frees just enough bbu toaccommodate incoming new calls. Figure 11 shows that theAJCAC scheme improves the system utilization by up to 20%of the NAJCAC scheme.


10/14


11109876543210

Call arrival rate

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

Normalized

averageutilization

NAJCACAJCAC

Figure 11: Impact of varying the call arrival rate on the normalizedaverage system utilization.

6. CONCLUSIONS

We propose adaptive bandwidth management and JCACscheme to enhance system utilization and connection-levelQoS in heterogeneous cellular networks. The adaptive JCACscheme improves average system utilization by adapting thebandwidth of calls based on current traffic condition and byuniformly distribute traffic load among the available RATs.The adaptive JCAC scheme guarantees the QoS requirementsof all accepted call and reduces both new call blocking prob-ability and handoffcall dropping probability in the heteroge-neous wireless networks. It prioritizes handoffcalls over new

calls by using different call rejection thresholds for new andhandoff calls. We develop a Markov chain model which en-ables us to derive new call blocking probability, handoffcalldropping probability, and average system utilization for theadaptive JCAC scheme. Performance of the adaptive JCACscheme is compared with that of nonadaptive JCAC schemein the same heterogeneous cellular network. Results showthat new call blocking probability and handoffcall droppingprobability can be significantly reduced by using the adap-tive JCAC scheme. Moreover, the adaptive JCAC scheme im-proves the system utilization by up to 20% of the nonadap-tive JCAC scheme.

REFERENCES

[1] A. Tolli, P. Hakalin, and H. Holma, Performance evaluationof common radio resource management (CRRM), in Proceed-ings of the IEEE International Conference on Communications(ICC 02), vol. 5, pp. 34293433, New York, NY, USA, April-May 2002.

[2] J. Perez-Romero, O. Sallent, R. Agust, and M. A. Daz-Guerra,Radio Resource Management Strategies in UMTS, JohnWiley &Sons, New York, NY, USA, 2005.

[3] R. Agust, O. Sallent, J. Perez-Romero, and L. Giupponi, Afuzzy-neural based approach for joint radio resource manage-ment in a beyond 3G framework, in Proceedings of the 1st In-ternational Conference on Quality of Service in Heterogeneous

Wired/Wireless Networks (QSHINE 04), pp. 216224, Dallas,Tex, USA, October 2004.

[4] R. M. Rao, C. Comaniciu, T. V. Lakshman, and H. V. Poor,Call admission control in wireless multimedia networks,IEEE Signal Processing Magazine, vol. 21, no. 5, pp. 5158,2004.

[5] M. H. Ahmed, Call admission control in wireless networks: a

comprehensive survey, IEEE Communications Surveys & Tu-torials, vol. 7, no. 1, pp. 4968, 2005.

[6] D. Karabudak, C.-C. Hung, and B. Bing, A call admissioncontrol scheme using genetic algorithms, in Proceedings of the

ACM Symposium on Applied Computing (SAC 04), pp. 11511158, Nicosia, Cyprus, March 2004.

[7] X. Gelabert, J. Perez-Romero, O. Sallent, and R. Agust, Onthe suitability of load balancing principles in heterogeneouswireless access networks, in Proceedings of the Wireless Per-sonal Multimedia Communications (WPMC 05), Aalborg,Denmark, September 2005.

[8] A. Pillekeit, F. Derakhshan, E. Jugl, and A. Mitschele-Thiel,Force-based load balancing in co-located UMTS/GSM net-works, in Proceedings of the 60th IEEE Vehicular Technology

Conference (VTC 04), vol. 6, pp. 44024406, Los Angeles,Calif, USA, September 2004.

[9] J. Perez-Romero, O. Sallent, and R. Agust, Policy-based ini-tial RAT selection algorithms in heterogeneous networks, inProceedings of the 7th IFIP International Conference on Mobileand Wireless Communications Networks (MWCN 05), Mar-rakesh, Morroco, September 2005.

[10] H. Holma and A. Toskala, WCDMA for UMTS, John Wiley &Sons, New York, NY, USA, 2nd edition, 2001.

[11] V. Pla, J. M. Gimenez-Guzmany, J. Martnez, and V. Casares-Giner, Optimal admission control using handover predictionin mobile cellular networks, in Proceedings of the 2nd Interna-tional Working Conference on Performance Modelling and Eval-uation of Heterogeneous Networks (HET-NETs 04), Ilkley, WestYorkshire, UK, July 2004.

[12] G. Kesidis, J. Walrand, and C.-S. Chang, Effective band-widths for multiclass Markov fluids and other ATM sources,IEEE/ACM Transactions on Networking, vol. 1, no. 4, pp. 424428, 1993.

[13] M. Gabowski, M. Stasiak, A. Wisniewski, and P.Zwierzykowski, Uplink blocking probability calculationfor cellular systems with WCDMA radio interface, finitesource population and differently loaded neighbouring cells,in Proceedings of the 11th Asia-Pacific Conference on Commu-nications (APCC 05), pp. 138142, Perth, Western Australia,October 2005.

[14] N. Nasser and H. Hassanein, Dynamic threshold-based calladmission framework for prioritized multimedia traffic inwireless cellular networks, in Proceedings of the IEEE Global

Telecommunications Conference (GLOBECOM 04), vol. 2, pp.644649, Dallas, Tex, USA, November-December 2004.

[15] Y. Guo and H. Chaskar, Class-based quality of service overair interfaces in 4G mobile networks, IEEE Communications

Magazine, vol. 40, no. 3, pp. 132137, 2002.[16] S.-P. Chung and J.-C. Lee, Performance analysis and over-

flowed traffic characterization in multiservice hierarchicalwireless networks, IEEE Transactions on Wireless Communi-cations, vol. 4, no. 3, pp. 904918, 2005.

[17] F. Houeto and S. Pierre, Characterization of jitter and admis-sion control in multiservice networks, IEEE CommunicationsLetters, vol. 8, no. 2, pp. 125127, 2004.

[18] J. Wang, Q.-A. Zeng, and D. P. Agrawal, Performance anal-ysis of a preemptive and priority reservation handoff scheme


11/14


12/14

:: Open Access ::

EURASIP Journal on

Wireless Communications and Networking

h t t p : / / w w w . h i n d a w i . c o m / j o u r n a l s / w c n /

Special Issue on

Wireless Access in Vehicular Environments

Call forPapers

Wireless access in vehicular environments (WAVE) technology comesinto sight as a state-of-the-art solution to contemporary vehicularcommunications, which is anticipated to be widely applied in the nearfuture to radically improve the transportation environment in aspectsof safety, intelligent management, and data exchange services. WAVEsystems will fundamentally smooth the progress of intelligent trans-portation systems (ITSs) by providing them with high-performancephysical platforms. WAVE systems will build upon the IEEE 802.11pstandard, which is still active and expected to be ratified in April,2009. Meanwhile, the VHF/UHF (700MHz) band vehicular commu-nication systems are attracting increasingly attention recently.

The fast varying and harsh vehicular environment brings about sev-eral fresh research topics on the study of WAVE systems and futurevehicular communication systems, which include physical layer chal-lenges associated with mobile channels, capacity evaluation, novelnetwork configuration, effective media access control (MAC) proto-cols, and robust routing and congestion control schemes.

The objective of this special section is to gather and circulate recentprogresses in this fast developing area of WAVE and future vehicularcommunication systems spanning from theoretical analysis to testbedsetup, and from physical/MAC layers enabling technology to networkprotocol. These research and implementation activities will be con-siderably helpful to the design of WAVE and future vehicular commu-nications systems by removing major technical barriers and present-ing theoretical guidance. This special issue will cover, but not limitedto, the following main topics:

700MHz/5.8GHz vehicle-to-vehicle (V2V)/vehicle-to-infra-structure (V2I) single-input single-output (SISO)/MIMOchannel measurement and modeling, channel spatial and tem-poral characteristics exploration

Doppler shift study, evaluation and estimate, time and fre-quency synchronizations, channel estimate and prediction Utilization of MIMO, space-time coding, smart antenna, adap-

tive modulation and coding Performance study and capacity analysis of V2V and V2I com-

munications operating over both 5.8GHz and 700Mz Software radio, cognitive radio, and dynamic spectrum access

technologies applied to WAVE and future vehicular communi-cation systems

Mesh network and other novel network configurations for ve-hicular networks

Efficient MAC protocols development

Routing algorithms and congestion control schemes for bothreal-time traffic warning message broadcasting and high-speeddata exchange

Cross-layer design and optimization Testbed or prototype activities

Authors should follow the EURASIP Journal on Wireless Com-munications and Networking manuscript format described at the

journal site http://www.hindawi.com/journals/wcn/. Prospective au-thors should submit an electronic copy of their complete manuscriptthrough the EURASIP Journal on Wireless Communications andNetworkings Manuscript Tracking System at http://mts.hindawi.com/, according to the following timetable.

Manuscript Due April 1, 2008

First Round of Reviews July 1, 2008

Publication Date October 1, 2008

Guest Editors

WeidongXiang, University of Michigan, Dearborn, USA;[email protected]

Javier Gozalvez, University Miguel Hernandez, Spain;[email protected]

Zhisheng Niu, Tsinghua University, China;[email protected]

OnurAltintas, Toyota InfoTechnology Center, Co. Ltd, Tokyo,Japan; [email protected]

EylemEkici, Ohio State University, USA; [email protected]

HindawiPublishingCorporation

410 Park Avenue, 15th Floor, #287 pmb, New York, NY 10022, USA

INDAWI
http://www.hindawi.com/journals/wcn/http://mts.hindawi.com/http://mts.hindawi.com/http://mts.hindawi.com/mailto:[email protected]:[email protected]://mts.hindawi.com/http://mts.hindawi.com/http://www.hindawi.com/journals/wcn/


13/14

Steering Committee:Prof. Z Ghassemlooy (Northumbria Univ., UK) ChairmanProf. A C Boucouvalas (Univ. of Peloponnese, Greece) Symp. Secr.Prof. R A Carrasco (Newcastle Univ. UK)Dr. Erich Leitgeb (Graz Univ. Austria) Local Organiser

Local Organising Committee:Dr. Erich Leitgeb (Local Organising Committee Chair)Prof. O. Koudelka (Professor)Dr U. BirnbacherJ. Ritsch and P. Reichel (OVE)Dr J. WolkerstorferP. Schrotter

National/International Technical Committee:Dr M Ali (Oxford Brookes Univ. UK)Prof. E Babulak (American Univ. Girne, North Cyprus)Prof. P Ball (Oxford Brookes Univ. UK)Dr D Benhaddou (University of Houston, USA)Prof. J L Bihan (Ecole Nation. d'Ingen. de Brest, France)Dr K E Brown (Heriot-Watt Univ. UK)Prof. R A Carrasco (Newcastle Univ. UK)Dr J B Carruthers (Boston University, USA)Prof. F Castanie (INPT, France)Dr H Castel (Inst. Natio. D. Tlcommun., France)Dr L Chao (Nanyang Tech. Univ. Singapore)Prof. R J Clarke (Heriot-Watt Univ. UK)Dr M Connelly (Univ. of Limerick, Ireland)Prof. A Constantinides (Imperial College, UK)Prof. D Dietrich, (Vienna Univ. of Tech., Austria) IEEEAustria Section ChairDr D Dimitrov (Tech. Univ. of Sofia, Bulgaria)Dr S Dlay (Newcastle Univ. UK)Prof. Fu, Shan (Shanghai Jiaotong Univ., China)Dr. M Gebhart, (NXP, Gratkorn Austria)Dr M Giurgiu (Univ. of Cluj-Napoca, Romania)Dr. M Glabowski, (Poznan Univ. of Tech., Poland)Prof. Ch Grimm, (Vienna Univ. of Tech., Austria)Dr A. Inoue (NTT, Japan)Prof. L Izzo (Univ. of Napoli, Italy)Prof. G Kandus (IJS Ljubljana, Slovenia)Prof. K Kawashima (Tokyo Univ. TUAT, Japan)Prof. C Knutson (Brigham Young Univ., USA)Prof. G Kubin, (Graz Univ. of Tech., AustriaProf. K Liu (Reading Univ. UK)Dr M D Logothetis (Univ. of Patras, Greece)Prof. E Lutz, (DLR, Oberpfaffenhofen, Germany)Prof. M Matijasevic (FER Zagreb, Croatia)Prof. B Mikac (FER Zagreb, Croatia)Dr. W P Ng (Northumbria University, UK)Dr T Ohtsuki (Tokyo Univ. of Sci., Japan)Prof. F Ozek (Ankara Univ., Turkey)Prof. R Penty (Cambridge Univ. UK)

Prof. W Pribyl, (Graz Univ. of Tech. Austria)Prof. J A Robinson (Univ. of York, UK)Dr D Roviras (INPT, France)Prof. M Rupp, (Vienna Univ. of Tech., Austria)Dr. S. Sheikh Muhammad (Univ. of Eng. & Tech.,Pakistan)Dr S Shioda (Chiba Univ. Japan)Dr J Sodha (Univ. of West Indies, Barbados, W. Indies)Dr I Soto (Santiago Univ. Chile)Dr U Speidel (Univ. of Auckland, Newzeland)Prof. M Stasiak (Poznan Univ. Poland)Dr L Stergioulas (Brunel Uni. UK)Prof. M Theologou (Nation. Tech. Univ. Athens, Greece)Prof. R. Vijaya (Indian Inst. of Tech., Bombay, India)Dr I Viniotis (N.Caroline State Univ.USA)Dr V Vitsas (TEI of Thessaloniki, Greece)Prof. R Wei, (Graz Univ. of Tech., Austria)Dr P. Xiao (Queens Univ. UK)Prof. M N Zervas Southam ton Univ. UK

Following the success of the last event, and after 10 years, the CSNDSP steeringcommittee decided to hold the next event at the Graz University, Austria. Graz wasthe 2003 cultural capital of Europe. CSNDSP, a biannual conference, started inUK ten years ago and in 2006 it was hold for the first time outside UK inPatras/Greece. CSNDSP has now been recognised as a forum for the exchangeof ideas among engineers, scientists and young researchers from all over theworld on advances in communication systems, communications networks, digitalsignal processing and other related areas and to provide a focus for futureresearch and developments. The organising committee invites you to submitoriginal high quality papers addressing research topics of interest for presentationat the conference and inclusion in the symposium proceedings.

Papers are solicited from, but not limited to the following topics: Adaptive signal processing ATM systems and networks

Chip design for Communications Communication theory Coding and error control Communication protocols Communications for disaster

management Crosslayer design DSP algorithms and applications E-commerce and e-learning

applications Intelligent systems/networks Internet communications High performance networks Mobile communications,

networks, mobile computing for e-commerce

Mobility management

Modulation and synchronisation Modelling and simulation

techniques Multimedia communications and

broadband services Microwave Communications

New techniques in RF-design andmodelling

Network management &operation

Optical communications Optical MEMS for lightwave

networks RF/Optical wireless

communications Photonic Network Quality of service, reliability and

performance modelling Radio, satellite and space

communications RFID & near field

communications Satellite & space communications Speech technology Signal processing for storage

Teletraffic models and trafficengineering VLSI for communications and

DSP Wireless LANs and ad hoc

networks 3G/4G network evolution Any other related topics

Papers may be presented in the form ofOral presentation and/or Poster Contributions by MPhil/PhD research students are particularly encouraged.Submission Dates:

Electronic submission by e-mail to: [email protected] All papers will be refereed and published in the symposium

proceeding. Selected papers will be published in: The MediterraneanJournals of Computers and Networks and Electronics andCommunications, and possibly IET proceedings.

A number of travel grants and registration fee waivers will beoffered to the delegates.

Fees: 360.00 EURO, (Group Delegates of 3 persons: 760.00 Euro)Includes: A copy of the Symposium Proceedings, Lunches, and SymposiumDinner on the 24

thJuly.

CSNDSP08 General InformationContact: Dr. Erich Leitgeb -Local Organising Committee Chair

- Institute of Broadband Communications, Graz University of Technology, A-

8010 Graz, Inffeldg. 12, Tel.: ++43-316-873-7442, Fax.: ++43-316-463697.

Email: [email protected], Web-site: http://www.inw.tugraz.at/ibk

6th International Symposium onCOMMUNICATION SYSTEMS, NETWORKS AND DIGITAL SIGNAL PROCESSING (CSNDSP08)

23-25 July 2008, Graz University of Technology, Graz, Austriawww.csndsp.com

Hosted by: Institute of Broadband Communications, Department of Communications and Wave PropagationSponsored by:

The Mediterranean Journals:- Computer & Networks- Electronics & Communications

IEEE CommunicationsChapter - UK/RI

First Call for Papers

Full Paper due: 27th

Jan. 2008

Notification of acceptance by: 1st

April 2008

Camera ready paper due: 5th

May 2008


14/14

CALL FOR PAPERS

3DTV CONFERENCE 2008THE TRUE VISION

CAPTURE, TRANSMISSION AND DISPLAY OF VIDEO

28-30 MAY 2008, HOTEL DEDEMAN, STANBUL, TURKEY

General Co-Chairs:

Uur Gdkbay, Bilkent University, TR

A. Aydn Alatan,Middle East Technical University, TR

Advisory Board:

Jrn Ostermann,Leibniz University of Hannover, DE

Aljoscha Smolic,Fraunhofer Gesellschaft, DE

A. Murat Tekalp, Ko University, TR

Levent Onural, Bilkent University, TR

John Watson, University of Aberdeen, UK

Thomas Sikora,Technische Universitaet Berlin, DE

Finance Chair:

Tark Reyhan, Bilkent University, TR

Publication Chair:

Tolga K. apn, Bilkent University, TR

Publicity Chairs:

Georgios Triantafyllidis,Centre for Research & Technology, GR

Gzde Bozda Akar,Middle East Technical University, TR

Industry Liaison:

Ismo Rakkolainen, FogScreen, FI

Matt Cowan, RealD, USA

American Liaison:

Kostas Danilidis,University of Pennsylvania, USA

Far East Liaison:

Kiyoharu Aizawa, University of Tokyo, JP

Special Sessions and

Tutorials Chairs:Philip Benzie, University of Aberdeen, UK

Atanas Gotchev,Tampere University of Technology, FI

Webmasters:

Engin Tretken,Middle East Technical University, TR

Aye Kkylmaz,

Bilkent University, TR

Following the conference of 2007, the second 3DTV Conference will be held in Istanbul, Turkey

in May, 2008. The aim of 3DTV-Con is to bring researchers from different locations together

and provide an opportunity for them to share research results, discuss the current problems

and exchange ideas.

The conference involves a wide range of research fields such as capturing 3D scenery, 3D image

processing, data transmission and 3D displays. You are cordially invited to attend 3DTV-Con

2008 and submit papers reporting your work related to the conference themes listed below.

Conference Topics

3D Capture and Processing:

- 3D time-varying scene capture technology

- Multi-camera recording- 3D photography algorithms

- Dense stereo and 3D reconstruction

- Synchronization and calibration of camera arrays

- 3D view registration

- Multi-view geometry and calibration

- Holographic camera techniques

- 3D motion analysis and tracking

- Surface modeling for 3D scenes

- Multi-view image and 3D data processing

- Integral imaging techniques

3D Visualization:

- 3D mesh representation

- Texture and point representation- Object-based representation and segmentation

- Volume representation

- 3D motion animation

- Stereoscopic display techniques

- Holographic display technology

- Reduced parallax systems

- Underlying optics and VLSI technology

- Projection and display technology for 3D videos

- Integral imaging techniques

- Human factors

3D Transmission:

- Systems, architecture and transmission in 3D

- 3D streaming

- Error-related issues and handling of 3D video

- Hologram compression

- Multi-view video coding

- 3D mesh compression

- Multiple description coding for 3D

- Signal processing for diffraction and holographic

3DTV

3D Applications:

- 3D imaging in virtual heritage and virtual archaeology

- 3D teleimmersion and remote collaboration

- Augmented reality and virtual environments

- 3D television, cinema, games and entertainment

- Underlying Technologies for 3DTV

- Medical and biomedical applications

- 3D content-based retrieval and recognition

- 3D watermarking

Paper Submission

Contributors are invited to submit full papers electronically using the online submission

interface, following the instructions at http://www.3dtv-con.org. Papers should be in Adobe

PDF format, written in English, with no more than four pages including figures, with a font

size of 11. Conference proceedings will be published online by IEEE Xplore.

Important Dates

Special sessions and tutorials proposals deadline: 14 December 2007

Regular paper submission deadline: 11 January 2008

Notification of paper acceptance: 29 February 2008

Camera-ready paper submission deadline: 21 March 2008

Conference: 28-30 May 2008

Sponsors

3DTV Network

of Excellence

Middle East

Technical

University

Bilkent

University

Institute of Electrical

and Electronics

Engineers

European

Association for

Signal and

Image Processing

MPEG Industry

Forum

adaptive bandwidth management and joint call

Documents