FACS-MP: A fuzzy admission control system with many priorities for wireless cellular networks and its performance evaluation

Abstract

Call Admission Control (CAC) is one of the resource management functions which regulates network access to ensure QoS provisioning. However, the decision for CAC is very challenging issue due to user mobility, limited radio spectrum and multimedia traffic characteristics. Several schemes have been proposed for CAC in wireless cellular networks. However, during to the complexity of CAC in wireless environment, many simplified models and assumptions are made. Some schemes consider that each mobile node will make hand-over to neighbouring cells with equal probability, which may be not accurate in general. For this reason, the intelligent and heuristic methods are needed. We proposed a fuzzy-based CAC system by considering the priority of the on-going connections. We called this system FACS-P. In FACS-P, as priority parameter, we considered only one parameter (service request). We extended our work by adding different priorities. We call this system FACS-MP. In this paper, we evaluate by simulations the performance of the proposed system. From the simulation results, we conclude that the FACS-MP makes a good differentiation of different services and has better behaviour than FACS-P.

Keywords

Cellular networks CAC fuzzy logic priorities intelligent algorithms

1 Introduction

As the demand for multimedia services over the air has been steadily increasing over the last few years, wireless multimedia networks have been a very active research area [10,19]. The Call Admission Control (CAC) is a good strategy to support various integrated services with certain Quality of Service (QoS) requirements. The CAC is a provisioning strategy that limits the number of connections into the network in order to reduce the network congestion and call dropping.

In cellular wireless networks when the mobile node moves from one cell to another one, the bandwidth must be requested in the new cell. During this process, the call may not be able to get a channel in the new cell due to the limited resource, which will lead to the call dropping. Thus, the new and handoff calls have to be treated differently in terms of resource allocation.

Since users are much more sensitive to call dropping than to call blocking, the handoff calls are assigned higher priority than new calls [23].

CAC techniques are required to guarantee QoS requirements for all traffic types. CAC is based on the knowledge of the statistical characteristics of on-going and arriving calls. The decision to accept an additional call involves the calculation or estimation of the consequences of the call acceptance on blocking and delay of itself and other incoming calls.

Several schemes have been proposed for CAC in wireless cellular networks. However, during the complexity of CAC in wireless environment, many simplified models and assumptions are made. Some schemes consider that each mobile node will make hand-over to neighbouring cells with equal probability, which may be not accurate in general. For this reason, the intelligent and heuristic methods are needed.

Use of intelligent methods based on Fuzzy Logic (FL), Neural Networks (NN) and Genetic Algorithms (GA) can prove to be efficient for traffic control in telecommunication networks [1,3–7,11,13,14,18,20–22].

In [2], in order to deal with CAC in wireless cellular networks, we proposed a CAC scheme based on FL. We implemented and evaluated the proposed system by comparing its performance with Shadow Cluster Concept (SCC) [15].

In [16,17], we proposed another FL-based CAC scheme by considering the priority of the on-going connections. However, as priority parameter, we considered only one parameter (service request). We called this system FACS-P. In [12], we extend our work by adding different priorities. We called this system: Fuzzy Admission Control System with Many Priorities (FACS-MP).

In this paper, we evaluate by simulations FACS-MP and compare its performance with FACS-P. From the simulation results, we found that the FACS-MP make a good differentiation of different services and has better performance than FACS-P.

The structure of this paper is as follows. In Section 2, we present the previous work. In Section 3, we introduce FACS-MP. In Section 4, we discuss the simulation results. Finally, some conclusions are given in Section 5.

2 Previous work

2.1 SCC

One of the previous work on CAC is SCC [15]. The fundamental idea of the SCC is that every mobile terminal with an active wireless connection exerts an influence upon the cells (and their BSs) in the vicinity of its current location and along its direction of travel. As an active mobile terminal travels to other cells, the region of influence also moves, following the active mobile terminal to its new location. The BSs (and their cells) currently being influenced are said to form a shadow cluster, because the region of influence follows the movements of the active mobile terminal like a shadow, as shown in Fig. 1.

The shadow is strongest near the active mobile terminal, and fades away depending on factors such as the distance to the mobile terminal, current call holding time and priority, bandwidth resources being used, and the mobile terminal’s trajectory and velocity. Because of these factors, the shape of a shadow cluster is usually not circular and can change over time. The center of a shadow cluster is not the geometric center of the area described by the shadow, but the cell where the mobile terminal is currently located. This cell is considered as the mobile terminal’s current home cell. A bordering neighbour is a cell that shares a common border with the shadow cluster’s center cell.

Conceptually, the number and “darkness” of the shadows covering a cell reflect the amount of resources that the cell’s BS needs to reserve in order to support the active mobile terminals currently in its own and in neighbouring cells. With the information provided by shadow clusters, BSs can determine, for each new call request, whether the request can be supported by the wireless network. In practice, a shadow cluster is a virtual message system where BSs share probabilistic information with their neighbours on the likelihood that their active mobile terminals will move to neighbour cells (while remaining active) in the near future. With the information provided by shadow clusters, BSs project future demands and reserve resources accordingly. BSs reserve resources by denying network access to new call requests, and by “waiting” for active users to end their calls.

The decision process for the acceptance of a new call request also involves a shadow cluster. Every new call request results in the implementation of a tentative shadow cluster. BSs exchange information on their new call requests, and decide, based on this and other information, which requests should be accepted and which requests should be denied.

Fig. 1.

SCC.

2.2 FL-based CAC

In our previous work, in order to deal with CAC in wireless cellular networks, we proposed a CAC scheme based on FL [2]. Conventional CAC schemes for wireless networks must consider some measured parameters to make the decision. However, in wireless networks due to the user mobility and varying of channel condition the measurement obtained are not accurate. Also, it is very difficult to obtain the complete statistics of the input traffic. Therefore, the CAC decision must be made based on the uncertain or inaccurate information. This is why we use FL. We implemented and evaluated the proposed system by comparing its performance with Shadow Cluster Concept (SCC) [15]. We showed that the proposed scheme could achieve a better prediction of the user behavior and a good admission decision compared with SCC.

In [16,17], we proposed another FL-based CAC system called FACS-P, which considered the following parameters for acceptance decision: user Speed ( $Sp$ ), user Angle ( $An$ ), Service request ( $Sr$ ), Correction value ( $Cv$ ), Required bandwidth ( $Rq$ ), Counter state ( $Cs$ ), Accept or Reject decision ( $A / R$ ), Differentiated service ( $Ds$ ), Real Time Counter ( $RTC$ ) and Non Real Time Counter ( $NRTC$ ). The structure of the FACS-P is shown in Fig. 2. In FACS-P, as the only priority parameter is considered the Service request ( $Sr$ ).

Fig. 2.

FACS-P model.

3 FACS-MP

In our previous work, we found that when using only one parameter for the priority, the FCAC-P system did not differentiate well different services. For this reason, in the new implemented system FACS-MP, we consider different priorities during the CAC decision.

The Fuzzy Logic Controller (FLC) is the main part of FACS-MP and its basic elements are shown in Fig. 3. They are the fuzzifier, inference engine, Fuzzy Rule Base (FRB) and defuzzifier. As membership functions, we use triangular and trapezoidal membership functions because they are suitable for real-time operation [8,9]. The membership functions are shown in Fig. 4.

Fig. 3.

FLC structure.

Fig. 4.

Triangular and trapezoidal membership functions.

The proposed FACS-MP considers the following parameters for acceptance decision: user Speed (S), user Angle (A), Correction value ( $Cv$ ), User priority ( $Up$ ), Service priority ( $Sp$ ), Connection priority ( $Cp$ ), Priority (P), Counter state ( $Cs$ ), Accept or Reject decision ( $A / R$ ), Differentiated service ( $Ds$ ), Real Time Counter ( $RTC$ ) and Non Real Time Counter ( $NRTC$ ). The structure of the proposed FACS-MP is shown in Fig. 5.

Fig. 5.

FACS-MP model.

3.1 FLC1 design

The input parameters for FLC1 are: user Speed (S) and user Angle (A), while the output linguistic parameter is Correction value ( $Cv$ ). The term sets of S, and A are defined respectively as: $\begin{array}{rcl} T (S) = {Slow, Middle, Fast} = {Sl, M, F}; \\ T (A) = {Back 1, Left 1, Left 2, Straight, Right 1, Right 2, Back 2} = {B 1, L 1, L 2, St, R 1, R 2, B 2} . \end{array}$

The membership functions for input parameters of FLC1 are defined as follows: $\begin{array}{rcl} μ_{Sl} (S) = g (S; {Sl}_{0}, {Sl}_{1}, {Sl}_{w 0}, {Sl}_{w 1}); \\ μ_{M} (S) = f (S; M_{0}, M_{w 0}, M_{w 1}); \\ μ_{F} (S) = g (S; F_{0}, F_{1}, F_{w 0}, F_{w 1}); \\ μ_{B 1} (A) = g (A; B 1_{0}, B 1_{1}, B 1_{w 0}, B 1_{w 1}); \\ μ_{L 1} (A) = f (A; L 1_{0}, L 1_{w 0}, L 1_{w 1}); \\ μ_{L 2} (A) = f (A; L 2_{0}, L 2_{w 0}, L 2_{w 1}); \\ μ_{St} (A) = f (A; {St}_{0}, {St}_{w 0}, {St}_{w 1}); \\ μ_{R 1} (A) = f (A; R 1_{0}, R 1_{w 0}, R 1_{w 1}); \\ μ_{R 2} (A) = f (A; R 2_{0}, R 2_{w 0}, R 2_{w 1}); \\ μ_{B 2} (A) = g (A; B 2_{0}, B 2_{1}, B 2_{w 0}, B 2_{w 1}) . \end{array}$ The small letters $w 0$ and $w 1$ mean left width and right width, respectively.

The term set of the output linguistic parameter $T (Cv)$ is defined as: ${Correction value 1, Correction value 2, \dots, Correction value 9} = {Cv 1, Cv 2, \dots, Cv 9} .$

The membership functions for the output parameter $Cv$ are defined as follows: $\begin{array}{rcl} μ_{Cv 1} (Cv) = g (Cv; Cv 1_{0}, Cv 1_{1}, Cv 1_{w 0}, Cv 1_{w 1}); \\ μ_{Cv 2} (Cv) = f (Cv; Cv 2_{0}, Cv 2_{w 0}, Cv 2_{w 1}); \\ μ_{Cv 3} (Cv) = f (Cv; Cv 3_{0}, Cv 3_{w 0}, Cv 3_{w 1}); \\ μ_{Cv 4} (Cv) = f (Cv; Cv 4_{0}, Cv 4_{w 0}, Cv 4_{w 1}); \\ μ_{Cv 5} (Cv) = f (Cv; Cv 5_{0}, Cv 5_{w 0}, Cv 5_{w 1}); \\ μ_{Cv 6} (Cv) = f (Cv; Cv 6_{0}, Cv 6_{w 0}, Cv 6_{w 1}); \\ μ_{Cv 7} (Cv) = f (Cv; Cv 7_{0}, Cv 7_{w 0}, Cv 7_{w 1}); \\ μ_{Cv 8} (Cv) = f (Cv; Cv 8_{0}, Cv 8_{w 0}, Cv 8_{w 1}); \\ μ_{Cv 9} (Cv) = g (Cv; Cv 9_{0}, Cv 9_{1}, Cv 9_{w 0}, Cv 9_{w 1}) . \end{array}$

The membership functions of FLC1 are shown in Fig. 6. The FRB forms a fuzzy set of dimensions

| T (S) | \times | T (A) |

, where

| T (x) |

is the number of terms on

T (x)

. The FRB1 shown in Table 1 has 21 rules. The control rules have the following form: IF “conditions” THEN “control action”.

Table 1
FRB1

Rule	Speed	Angle	$Cv$
1	$Sl$	$B 1$	$Cv 2$
2	$Sl$	$L 1$	$Cv 3$
3	$Sl$	$L 2$	$Cv 4$
4	$Sl$	S	$Cv 7$
5	$Sl$	$R 1$	$Cv 4$
6	$Sl$	$R 2$	$Cv 3$
7	$Sl$	$B 2$	$Cv 2$
8	M	$B 1$	$Cv 1$
9	M	$L 1$	$Cv 3$
10	M	$L 2$	$Cv 3$
11	M	$St$	$Cv 9$
12	M	$R 1$	$Cv 3$
13	M	$R 2$	$Cv 3$
14	M	$B 2$	$Cv 1$
15	$Fa$	$B 1$	$Cv 1$
16	$Fa$	$L 1$	$Cv 2$
17	$Fa$	$L 2$	$Cv 3$
18	$Fa$	S	$Cv 9$
19	$Fa$	$R 1$	$Cv 3$
20	$Fa$	$R 2$	$Cv 2$
21	$Fa$	$B 2$	$Cv 1$

Fig. 6.

FLC1 membership functions.

3.2 FLC2 design

The input parameters for FLC2 are: User priority ( $Up$ ), Connection priority ( $Cp$ ) and Service priority ( $Sp$ ), while the output linguistic parameter is user Priority (P). The term sets of $Up$ , $Cp$ and $Sp$ are defined respectively as: $\begin{array}{rcl} T (Up) = {Low priority, Middle priority, High priority} = {Lp, Mp, Hp}; \\ T (Cp) = {Short, Middle, Long} = {Sh, Mi, Lo}; \\ T (Sp) = {Text, Voice, Video} = {T, Vo, Vi} . \end{array}$

The term set of the output linguistic parameter $T (P)$ is defined as ${Priority 1, Priority 2, \dots, Priority 9} = {P 1, P 2, \dots, P 9} .$

The membership functions for the output parameter P are defined as follows: $\begin{array}{rcl} μ_{P 1} (P) = g (P; P 1_{0}, P 1_{1}, P 1_{w 0}, P 1_{w 1}); \\ μ_{P 2} (P) = f (P; P 2_{0}, P 2_{w 0}, P 2_{w 1}); \\ μ_{P 3} (P) = f (P; P 3_{0}, P 3_{w 0}, P 3_{w 1}); \\ μ_{P 4} (P) = f (P; P 4_{0}, P 4_{w 0}, P 4_{w 1}); \\ μ_{P 5} (P) = f (P; P 5_{0}, P 5_{w 0}, P 5_{w 1}); \\ μ_{P 6} (P) = f (P; P 6_{0}, P 6_{w 0}, P 6_{w 1}); \\ μ_{P 7} (P) = f (P; P 7_{0}, P 7_{w 0}, P 7_{w 1}); \\ μ_{P 8} (P) = f (P; P 8_{0}, P 8_{w 0}, P 8_{w 1}); \\ μ_{P 9} (P) = g (P; P 9_{0}, P 9_{1}, P 9_{w 0}, P 9_{w 1}) . \end{array}$

The membership functions of FLC2 are shown in Fig. 7. The FRB2 shown in Table 2 has 27 rules.

Table 2
FRB2

Rule	$Up$	$Cp$	$Sp$	P
1	$Lp$	$Sh$	T	$P 6$
2	$Lp$	$Sh$	$Vo$	$P 5$
3	$Lp$	$Sh$	$Vi$	$P 4$
4	$Lp$	$Mi$	T	$P 5$
5	$Lp$	$Mi$	$Vo$	$P 3$
6	$Lp$	$Mi$	$Vi$	$P 2$
7	$Lp$	$Lo$	T	$P 4$
8	$Lp$	$Lo$	$Vo$	$P 2$
9	$Lp$	$Lo$	$Vi$	$P 1$
10	$Mp$	$Sh$	T	$P 5$
11	$Mp$	$Sh$	$Vo$	$P 6$
12	$Mp$	$Sh$	$Vi$	$P 5$
13	$Mp$	$Mi$	T	$P 4$
14	$Mp$	$Mi$	$Vo$	$P 5$
15	$Mp$	$Mi$	$Vi$	$P 4$
16	$Mp$	$Lo$	T	$P 3$
17	$Mp$	$Lo$	$Vo$	$P 4$
18	$Mp$	$Lo$	$Vi$	$P 3$
19	$Hp$	$Sh$	T	$P 8$
20	$Hp$	$Sh$	$Vo$	$P 9$
21	$Hp$	$Sh$	$Vi$	$P 9$
22	$Hp$	$Mi$	T	$P 7$
23	$Hp$	$Mi$	$Vo$	$P 8$
24	$Hp$	$Mi$	$Vi$	$P 9$
25	$Hp$	$Lo$	T	$P 6$
26	$Hp$	$Lo$	$Vo$	$P 7$
27	$Hp$	$Lo$	$Vi$	$P 8$

Fig. 7.

FLC2 membership functions.

3.3 FLC3 design

The input parameters for FLC3 are: the output parameter of the FLC1 ( $Cv$ ), user Priority (P) and the Counter state ( $Cs$ ), which shows the capacity of the system. While, the output linguistic parameter is the Accept/Reject decision ( $A / R$ ).

The term sets of $Cv$ , P and $Cs$ are defined as: $\begin{array}{rcl} T (Cv) = {Bad, Normal, Good} = {B, N, G}; \\ T (P) = {Low priority, Middle priority, High priority} = {Lwp, Mdp, Hgp}; \\ T (Cs) = {Small, Middle, Full} = {Sm, Md, Fu} . \end{array}$

In order to have a soft admission decision, for the output linguistic parameter ( $A / R$ ), we considered not only “accept” and “reject” but also “weak accept”, “weak reject” and “not accept not reject” for the accept/reject decision. The membership functions for input and output linguistic parameters of FLC3 are shown in Fig. 8. The FRB3 shown in Table 3 has 27 rules.

The membership functions for input parameters of FLC3 are defined as follows: $\begin{array}{rcl} μ_{B} (Cv) = f (Cv; B_{0}, B_{w 0}, B_{w 1}); \\ μ_{N} (Cv) = f (Cv; N_{0}, N_{w 0}, N_{w 1}); \\ μ_{G} (Cv) = f (Cv, G_{0}, G_{w 0}, G_{w 1}); \\ μ_{Lwp} (P) = f (P; {Lwp}_{0}, {Lwp}_{w 0}, {Lwp}_{w 1}); \\ μ_{Mdp} (P) = f (P; {Mdp}_{0}, {Mdp}_{w 0}, {Mdp}_{w 1}); \\ μ_{Hgp} (P) = f (P; {Hgp}_{0}, {Hgp}_{w 0}, {Hgp}_{w 1}); \\ μ_{Sm} (Cs) = f (Cs; {Sm}_{0}, {Sm}_{w 0}, {Sm}_{w 1}); \\ μ_{Md} (Cs) = f (Cs; {Md}_{0}, {Md}_{w 0}, {Md}_{w 1}); \\ μ_{Fu} (Cs) = f (Cs; {Fu}_{0}, {Fu}_{w 0}, {Fu}_{w 1}) . \end{array}$

The term set of the output linguistic parameter

T (A / R)

is defined as

{Reject, Weak Reject, Not Reject Not Accept, Weak Accept, Accept}

. We write for short as

{R, WR, NRNA, WA, A}

. The membership functions for the output parameter

A / R

are defined as follows:

\begin{array}{rcl} μ_{R} (A / R) = g (A / R; R_{0}, R_{1}, R_{w 0}, R_{w 1}); \\ μ_{WR} (A / R) = f (A / R; {WR}_{0}, {WR}_{w 0}, {WR}_{w 1}); \\ μ_{NRNA} (A / R) = f (A / R; {NRNA}_{0}, {NRNA}_{w 0}, {NRNA}_{w 1}); \\ μ_{WA} (A / R) = f (A / R; {WA}_{0}, {WA}_{w 0}, {WA}_{w 1}); \\ μ_{A} (A / R) = g (A / R; A_{0}, A_{1}, A_{w 0}, A_{w 1}) . \end{array}

Table 3
FRB3

Rule	$Cv$	P	$Cs$	$A / B$
1	B	$Lwp$	$Sm$	$WA$
2	B	$Lwp$	$Md$	$NRNA$
3	B	$Lwp$	$Fu$	$WR$
4	B	$Mdp$	$Sm$	A
5	B	$Mdp$	$Md$	$NRNA$
6	B	$Mdp$	$Fu$	$WR$
7	B	$Hgp$	$Sm$	A
8	B	$Hgp$	$Md$	$NRNA$
9	B	$Hgp$	$Fu$	$NRNA$
10	N	$Lwp$	$Sm$	$WA$
11	N	$Lwp$	$Md$	$NRNA$
12	N	$Lwp$	$Fu$	$NRMA$
13	N	$Mdp$	$Sm$	A
14	N	$Mdp$	$Md$	$NRNA$
15	N	$Mdp$	$Fu$	$NRNA$
16	N	$Hgp$	$Sm$	A
17	N	$Hgp$	$Md$	$NRNA$
18	N	$Hgp$	$Fu$	$NRNA$
19	G	$Lwp$	$Sm$	A
20	G	$Lwp$	$Md$	A
21	G	$Lwp$	$Fu$	R
22	G	$Mdp$	$Sm$	A
23	G	$Mdp$	$Md$	A
24	G	$Mdp$	$Fu$	$WR$
25	G	$Hgp$	$Sm$	A
26	G	$Hgp$	$Md$	A
27	G	$Hgp$	$Fu$	$NRMA$

Fig. 8.

FLC3 membership functions.

4 Simulation results

We considered the following parameters for simulations:

the user speed was from 0 to 120 km/h,

the user direction was changed from −180 degree to +180 degree,

the requested size was 1, 5 and 10 Bandwidth Units (BU) for text, voice and video, respectively,

the bandwidth of the BS was considered 40 BU.

In our previous work, we presented the performance of the previous FL-based CAC systems. We have shown that FACS-P has better behaviour than previous FL-based systems. However, by using only one priority parameter, the FACS-P system cannot differentiate well different services as shown in Fig. 9.

In Figs 10 and 11, we show the performance of FACS-MP. As can be seen by these figures, the proposed system shows a good behaviour, by making a good differentiation for different services and different priorities.

In Fig. 12, we show the performance comparison between FACS-MP and FACS-P. When the number of requesting connections is small, the percentage of accepted calls for FACS-P and FACS-MP is almost the same. However, when the number of requesting connections is larger than 45, the FACS-MP system accepts more number of connections.

In Figs 13, 14 and 15, we show the performance of priority algorithm in FACS-MP. In Fig. 13 is shown the relation between the priority and $Up$ , $Sp$ when $Cp$ is 0.5. With the increase of $Up$ and $Sp$ , the priority parameter is increased.

In Fig. 14, we increased the value of $Cp$ from 0.5 to 4.5. We see that when the $Cp$ is increased, the priority parameter increases, too. In Fig. 15, the $Cp$ parameter is increased to 9 units. We see that in this case, when a user has a long time connection call, the priority increases much more than the previous two cases. This is because, for short connection calls they can be covered by a present BS, but for long time call connections they may need to make handover to other BSs, so they get higher priority.

Fig. 9.

Performance of FACS-P for different services.

5 Conclusions

In this paper, we presented the FACS-MP system design and its performance evaluation, We evaluated the FACS-MP system performance for different services and priorities and compared with FACS-P.

From the simulations results, we conclude that the proposed FACS-MP makes a good differentiation for different services and priorities. Also, when the number of requesting connections is larger than 45, the FACS-MP system accepts more connections than FACS-P.

In the future, we would like to evaluate the proposed FACS-MP by extensive simulations.

Fig. 10.

Performance of FACS-MP for different services.

Fig. 11.

Performance of FACS-MP for different priorities.

Fig. 12.

Performance comparison between FACS-MP and FACS-P.

Fig. 13.

Performance of priority algorithm when $Cp = 0.5$ . (Colors are visible in the online version of the article; https://dx-doi-org.web.bisu.edu.cn/10.3233/JHS-150504.)

Fig. 14.

Performance of priority algorithm when $Cp = 4.5$ . (Colors are visible in the online version of the article; https://dx-doi-org.web.bisu.edu.cn/10.3233/JHS-150504.)

Fig. 15.

Performance of priority algorithm when $Cp = 9$ . (Colors are visible in the online version of the article; https://dx-doi-org.web.bisu.edu.cn/10.3233/JHS-150504.)

References

[1]

Barolli,

Sakamoto,

Oda,

Spaho,

Barolli and

Xhafa, Performance evaluation of WMN-GA system for different settings of population size and number of generations, Human-centric Computing and Information Sciences (HCIS) 4(5) (2014), 1–14.

[2]

Barolli, An intelligent call admission control system for wireless cellular networks based on fuzzy logic, Journal of Mobile Multimedia (JMM) 3(4) (2007), 331–346.

[3]

Barolli, A speed-aware handover system for wireless cellular networks based on fuzzy logic, Mobile Information Systems (MIS) 4(1) (2008), 1–12.

[4]

Barolli,

Anno,

Xhafa,

Durresi and

Koyama, A context-aware fuzzy-based handover system for wireless cellular networks and its performance evaluation, Journal of Mobile Multimedia (JMM) 4(3,4) (2008), 241–258.

[5]

Barolli,

Koyama,

Suganuma and

Shiratori, GAMAN: A GA based QoS routing method for mobile ad-hoc networks, Journal of Interconnection Networks (JOIN) 4(3) (2003), 251–270.

[6]

Barolli,

Koyama,

Yamada,

Yokoyama,

Suganuma and

Shiratori, A fuzzy admission control scheme and its performance evaluation, IPSJ Journal 42(12) (2001), 3213–3221.

[7]

Bensaou,

S.T.C.

Lam,

H.-W.

Chu and

D.H.K.

Tsang, Estimation of the cell loss ratio in ATM networks with a fuzzy system and application to measurement-based call admission control, IEEE/ACM Transactions on Networking 5(4) (1997), 572–584.

[8]

Dubois,

Prade and

Yager (eds), Fuzzy Sets for Intelligent Systems, Morgan Kaufmann Publishers, 1993.

[9]

Esposito,

Ficco,

Palmieri and

Castiglione, Smart cloud storage service selection based on fuzzy logic, theory of evidence and game theory, IEEE Transactions on Computers (2015), DOI:10.1109/TC.2015.2389952.

10.

[10]

Fang and

Zhang, Call admission control schemes and performance analysis in wireless mobile networks, IEEE Transactions on Vehicular Technology 51(2) (2002), 371–382.

11.

[11]

Habib (ed.), IEEE Communications Magazine 33(10) (1995), 76–112. (Special Issue: Neurocomputing in High-Speed Networks.)

12.

[12]

Ikebata,

Mino,

Barolli,

Koyama,

Durresi and

Xhafa, A fuzzy-based CAC scheme for wireless cellular networks considering different priorities, in: Proceedings of International Conference on Network-Based Information Systems (NBiS-2010), Japan, September 2010, 2010, pp. 187–194.

13.

[13]

Ikeda,

Barolli,

Koyama,

Durresi,

De Marco and

Iwashige, Performance evaluation of an intelligent CAC and routing framework for multimedia applications in broadband networks, Journal of Computer and System Science (JCSS) 72(7) (2006), 1183–1200.

14.

[14]

Ikeda,

Honda,

Oda,

Sakamoto and

Barolli, Analysis of WMN-GA simulation results: WMN performance considering stationary and mobile scenarios, in: The 28th International Conference on Advanced Information Networking and Applications (AINA 2014), Victoria, Canada, 13–16 May 2014, 2014, pp. 337–342.

15.

[15]

D.A.

Levine,

I.F.

Akyildiz and

Naghshineh, A resource estimation and call admission algorithm for shadow cluster concept, IEEE/ACM Transactions on Networking 5(1) (1997), 1–12.

16.

[16]

Mino,

Barolli,

Durresi,

Xhafa and

Koyama, A fuzzy-based call admission control scheme for wireless cellular networks considering priority of on-going connections, in: Proceedings of ICDCS-2009 Workshops, MNSA-2009, Montreal, Canada, June 2009, 2009, pp. 380–387.

17.

[17]

Mino,

Barolli,

Xhafa,

Durresi and

Koyama, Performance evaluation of a fuzzy-based CAC scheme for wireless cellular networks: A case study considering priority of on-going connections, International Journal of Business Intelligence and Data Mining 5(3) (2010), 269–284.

18.

[18]

Oda,

Sakamoto,

Barolli,

Spaho,

Barolli and

Xhafa, Effect of different grid shapes in wireless mesh network genetic algorithm system, Journal of Web and Grid 10(4) (2014), 371–395.

19.

[19]

Palmieri,

Fiore and

Castiglione, Automatic security assessment for next generation wireless mobile networks, Mobile Information Systems (MIS) 7(3) (2011), 217–239.

20.

[20]

Sakamoto,

Kulla,

Oda,

Ikeda,

Barolli and

Xhafa, Performance evaluation considering iterations per phase and SA temperature in WMN-SA system, Mobile Information Systems (MIS) 10(3) (2013), 321–330.

21.

[21]

Sakamoto,

Kulla,

Oda,

Ikeda,

Barolli and

Xhafa, A comparison study of simulated annealing and genetic algorithm for node placement problem in wireless mesh networks, Journal of Mobile Multimedia (JMM) 9(1,2) (2013), 101–110.

22.

[22]

Sakamoto,

Kulla,

Oda,

Ikeda,

Barolli and

Xhafa, A simulation system for wireless mesh networks based on simulated annealing: Performance evaluation for different instances and starting temperature values, International Journal of Space-Based and Situated Computing 4(3,4) (2014), 209–216.

23.

[23]

Wang,

Wang and

A.A.

Nilsson, Energy-efficient bandwidth allocation in wireless networks: Algorithms, analysis, and simulations, IEEE Transactions on Wireless Communications 5(5) (2006), 1103–1114.

FACS-MP: A fuzzy admission control system with many priorities for wireless cellular networks and its performance evaluation

Abstract

Keywords

1 Introduction

2 Previous work

2.1 SCC

Table 1 FRB1

Table 2 FRB2

Table 3 FRB3

References

Table 1
FRB1

Table 2
FRB2

Table 3
FRB3