To achieve the stated objective, separate buffer for each arrival cells for Class j (λj) is queued in First-In-First-Out
individual class is used to accommodate different traffic (FIFO) buffer while waiting to be served.
classes. This will eliminate the head of line (HoL) of
different classes blocking effects. For different port, Virtual Low priority
Output Queue (VOQ) [6] [7] is used in order to eliminate
the HoL of different destination port blocking. HoL
In this paper, we propose adaptive hybrid scheduling
method by combining two thresholds setting and able to
High priority
adjust the priority level of scheduling according to the mean HoL Sch
queue length of the delay sensitive class. In addition, the
scheduling technique is non-preemptive, which is more Figure 2. Head of Line Scheduler.
efficient and less complex than preemptive approaches due
to reducing the overhead needed for switching among cells. At each time slot, the switch attempts to serve the cells at
Head of Line (HoL) of each input queue as shown in Figure
2. Proposed Model 2. In the case when there are cells from different classes are
The proposed multi-class switch and the scheduling waiting at HoL, the HoL scheduler (HoL Sch) will select the
technique used in controlling the flows of the cells in the cell with high priority to be served. The losing cells in the
contention must wait in the queue. The numbers of queue
switch is described below.
cells will increase when there are new incoming cells to the
queue.
2.1 System model The threshold setting is introduced in order to give some
The proposed multi-class switch architecture with N privileges to cells in lower priority class. The threshold
ports serving C-1 classes of traffics is shown in Figure 1. parameter used in this architecture are the number of queue
Priority switch is used to forward delay sensitive packet cell, Nbj and the probability of serving low priority cell,
(class C-1) faster than loss sensitive packet (class 0). P TSCj; j = 0, 1, 2, …., C-1. Nbj parameter is chosen because
of the limited buffer size available in practical design. The
need to adjust the Nbj parameter is necessary to reduce the
packet loss due to buffer full. This parameter is adjusted
based on the size of buffer used to store cells in Class j. P TSCj
is the probability to serve the Class j when the Nbj parameter
threshold is met. The P TSCj parameter is chosen in order to
control the variation of delay between high priority and low
priority cells based on the high priority QoS requirements.
This is necessary in order to achieve better performance for
Class j cells. In the case where both threshold values are
met, the switch will select the cell from the class which the
threshold is triggered even in the present of higher priority
cells.
2.1 Simulation model
A simulation model is developed to simulate the
performance of the proposed switch under dual thresholds
setting. In this simulation, the architecture uses 16 x 16
Figure 1. Multi-class switch architecture switch with two classes for every input port. Class 0 is used
to classify the low priority buffer for non-real-time data. At
The delay requirement for class j cells is defined by Dj the same time, Class 1 represents high priority buffer for
where j = 0, 1, 2… C-1 and D0 > D1 > D2 >……> DC-1. In real-time data.
other words, the delay requirement for Class 1 is more The design of the switch architecture for input queue
stringent than Class 0 for system shown in Figure 1. Thus, multi-class switch is using separate buffer for each class.
cells that queue in Class 0 have the lowest priority and cells Arrival cells are stored in different FIFO based on their
that queue in Class C-1 have the highest priority. These classes. The HoL scheduler will choose one cells from HoL
delay requirements are set based on the QoS requirement for FIFO classes at every port to be forwarded to switch fabric.
different type of applications. The cells will contend with each other to gain access for
Time slot is used to represent the time of one cell arrival departure.
at the input port or cell departure at the output port. The The proposed switch operates in time slotted transmission
period of time slot Ts, where s = 0, 1, 2, … is set to be equal to process each cells. Each time slot consists of three phases
to the time to process a single cell when the server is idle. which are arrival, scheduler and departure.
The class j cells arrive at the input port in every time slot In the arrival phase, the incoming packet are segmented
according to Bernoulli distributions with mean λj. The cell into fixed size packet called cells and are aligned for
is classified based on its delay requirement. In this synchronization. The number of maximum cells, P max which
architecture the class of the cell is stored in the header. An is generated in one time slot depends on the traffic load, λ
(IJCNS) International Journal of Computer and Network Security, 23
Vol. 2, No. 8, August 2010
In the class segment, bit ‘1’ indicates that the cell belongs
to the delay sensitive class and bit ‘0’ is for the loss sensitive
class. This bit is classified based on the type of packet
received and the QoS requirement for its applications. Each
generated cell is classified either to the delay sensitive class
or the loss sensitive class based on the traffic type. High
priority cell is tagged with 1 and low priority cell with 0. Figure 6. HoL scheduler design
Then, the cell is sent to FIFO waiting to be served. The Figure 7 illustrates the cell flow at the HoL scheduler under
Class 1 cell is sent to FIFO_H1 and Class 0 cell to three situations which are based on threshold condition and
FIFO_L1. The HoL for Class 1 and Class 0 must wait until destination address of class 1 and class 0.
it is served. In general, the HoL scheduler will choose the
cell in Class 1 since it has the high priority cells. In the case
when both threshold values are achieved, the HoL scheduler √ Different
will choose the Class 0 cells even in a presence of Class 1 √ destination
cells. address
The pseudo code for the input buffer with HoL output Class1 Same
scheduling is shown in Figure 5. When HoL for class 1 is √ destination
not empty and the threshold value setting is applicable for
HoL
× address, Nb &
PTSC0 = 0
both parameters, the HoL output will choose HoL packet Class0 WAIT
from class 0 instead of class 1. scheduler
WAIT Same
If (HoL class1 not empty) × destination
If (Nb0 > TN) && (PTSC0 >TP) √ address, Nb &
HoL_out = HoL class0; PTSC0 = 1
else
HoL_out = HoL class1;
else Figure 7. Example of success and failure of cell flow at
HoL_ out = HoL class0; HoL Scheduler
Class 0
λ0 HoL Sch 1
16 X 16
Nonblocking
Switch
λ1 Class 1 Fabric
Hardware design is developed to evaluate the performance equal distribution of λ1 and λ0.
of the proposed architecture. Figure 11 shows the hardware Total Mean
Delay
timing simulation of incoming and outgoing cells in multi- (time slot)
class switch. It can be seen that there are new incoming cell (i) Class 0 (PQ)
(ii) Class 0 (adaptive threshold)
at every time slots. After the cells have been processed, only (iii) Class 0 (WFQ) (i)
10
3 (iv) Class 1 (WFQ) (iii)
the successful cell is allowed to depart. The rest of the cells (v) Class 1 (adaptive threshold) (iv)
(vi) Class 1 (PQ) (ii)
must waiting for their turns. The output cell timing is used
to calculate the total means delay for Class 1 and Class 0. 10
2
(v)
1
10
(vi)
0 0.9 1
10
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Ratio (r)
0.35
(iv)
(v)
Class 0 (adaptive threshold)
Class 0 (WFQ) (iv)
4. Conclusions
(vi) Class 0 (PQ)
(v) In this paper, the multi-class switch with adaptive dual
0.30 (vi)
threshold is proposed to optimize the performance of class 0
0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1
Load
traffic without affecting the QoS requirement for class 1.
The simulation results show that the mean delay and
Figure 12. Throughput in multi-class switch with equal
throughput using adaptive threshold is better than WFQ. In
distribution of λ1 and λ0. adaptive threshold, the P TSC0 and Nb parameter are adjusted
Total Mean
Delay automatically based on condition of the traffic load. By
(time slot)
adjusting these parameters adaptively the optimum of mean
(i) Class 0 (PQ)
(ii) Class 0 (adaptive threshold) (i) delay and throughput for class 0 can be achieved without
(iii) Class 0 (WFQ) (ii)
10 2 (iv) Class 1 (WFQ)
(iii)
degrading the QoS requirement for class 1.
(v) Class 1 (adaptive threshold) (iv)
(vi) Class 1 (PQ)
10
1 References
[1] Choi, J. S. and C. K. Un, "Delay Performance of an
(v)
10 0 Input Queueing Packet Switch with Two Priority
(vi) Classes". Communications, IEE Proceedings- Vol.
10
-1 145 (3): pp. 141-144, 1998.
0. 0.65 0. 0.75 0.
Load
[2] Warde, W. and P. A. Ivey, "Input Queueing Multicast
Atm Packet Switch with Two Priority Classes Using a
Figure 13. Total mean delay in multi-class switch with
26 (IJCNS) International Journal of Computer and Network Security,
Vol. 2, No. 8, August 2010