Method and system for traffic flow control in a communication network
View Patent ↗A method for controlling traffic flow in a communication network compliant with a Multilevel Precedence and Preemption (MLPP) scheme is described. The method selectively modifies parameters of TCP connections for transmissions to meet speed of service (SoS) requirements and establishes the connections in accordance with their levels of precedence. In applications, the method may be implemented in communication networks using High Assurance Internet Protocol Encryption (HAIPE) or an IP Security (IPSec) protocol.
1. A method for controlling traffic flow in an Internet Protocol (IP) based communication network, comprising software running on a computer acting as a network edge proxy in said network, said method comprising the steps of:
(a) associating a speed of service (SoS) with a level of precedence;
(b) selectively adjusting settings of Transmission Control Protocol (TCP) connections to meet the speed of service requirements for a particular level of precedence, said settings consisting of at least one of:
(b1) adjusting settings CWND adj for a congestion window;
(b2) adjusting RTO adj for a re-transmission time-out time interval using the equation:
RTO adj =min{( N×RTT e ), RTO max },
where N is a pre-determined constant in a range from 1 to 3, RTT e is an estimated transmission round trip time, and RTO max is a maximum re-transmission time-out interval determined using the equation:
∑
i
=
1
I
P
e
i
-
1
(
1
-
p
e
)
i
·
[
D
e
+
α
j
e
+
(
i
-
1
)
RTO
max
]
=
L
q
,
where p e is an estimated transmission loss ratio, D e is an estimated transmission transfer delay, j e is an estimated jitter delay α is a pre-determined constant, I is a number of re-transmissions, i is an integer, 1≦i≦I and L q is a transmission latency requirement; and
(b3) adjusting settings DL adj for time delays for sending acknowledgements of the transmissions;
(c) authorizing use of the TCP connections for transmissions having levels of precedence requiring the available SoS;
(d) establishing the TCP connections;
(e) monitoring compliance of the TCP connections with the SoS requirements of the transmission;
(f) dynamically controlling the TCP connections based on compliance with the required SoS; and
(g) repeating steps (b) through (g) until the transmission is complete.
2. The method of claim 1 , wherein the network is a wireless, wired, or fiber-optic communication network compliant with a Multilevel Precedence and Preemption (MLPP) scheme.
3. The method of claim 1 , wherein the network is compliant with at least one data encryption protocol.
4. The method of claim 3 , wherein the network uses High Assurance Internet Protocol Encryption (HAIPE) or an IP Security (IPSec) protocol.
5. The method of claim 1 , wherein the TCP connection is a connection between a transmitting edge proxy and a receiving edge proxy of a tactical network having an encrypted core.
6. The method of claim 5 , wherein steps (a)-(g) are performed at the transmitting network edge proxy.
7. The method of claim 1 , wherein step (b1) comprises:
determining the settings CWND adj using an equation:
CWND adj =max{ CWND currents ,CWND required },
where CWND current is a current setting for the congestion window, CWND required =(RTT e ×MSS)/L q , RTT e is an estimated transmission round trip time, MSS is a maximum size of a transmission unit, and L q is a transmission latency requirement.
8. The method of claim 1 , wherein RTO max is determined using an equation:
RTO max =( L q −D e −αj e )/( I+ 1),
where L q is a transmission latency requirement, D e is a transmission transfer delay, α is a pre-determined constant, j e is a jitter delay, and I is a number re-transmissions.
9. The method of claim 1 , wherein step (b3) comprises:
adjusting the settings DL adj as follows, where RTT e is an estimated transmission round trip time and T m is a pre-determined time interval:
TRAFFIC
CONDITIONS
TIMING FOR SENDING ACKNOWLEDGEMENT
RTT e ≦ T m
Sending acknowledgement with time delay DL adj = 0.
RTT e > T m
Sending acknowledgement with time delay
DL adj = RTT e − T m .
10. The method of claim 1 , wherein step (c) comprises:
authorizing the TCP connections meeting both of the following conditions:
RTO max >βRTT e , and
RWND>CWND required ,
where RTO max is a maximum re-transmission time-out interval; β is a pre-determined constant, RTT e is an estimated transmission round trip time, RWND is a maximum size of a congestion window at a destination of the transmission, and CWND required =(RTT e ×MSS)/L q , RTT e is an estimated transmission round trip time, MSS is a maximum size of a transmission unit, and L q is a transmission latency requirement.
11. The method of claim 10 , wherein RTO max is determined using an equation:
∑
i
=
1
I
P
e
i
-
1
(
1
-
p
e
)
i
·
[
D
e
+
α
j
e
+
(
i
-
1
)
RTO
max
]
=
L
q
,
where p e is an estimated transmission loss ratio, D e is an estimated transmission transfer delay, j e is an estimated jitter delay, α is a pre-determined constant, I is a number of re-transmissions, i is an integer, 1≦I≦I, and L q is a transmission latency requirement.
12. The method of claim 10 , therein RTO max is determined using an equation:
RTO max =( L q −D e −αj e )/( I+ 1),
where L q is a transmission latency requirement, D e is a transmission transfer delay, α is a pre-determined constant, j e is a jitter delay, and I is a number re-transmissions.
13. The method of claim 1 , wherein step (d) further comprises:
dynamically controlling the TCP connections as follows, wherein Level of Precedence 1 (LP1)<Level of Precedence 2 (LP2)<Level of Precedence 3 (LP3):
TRAFFIC CONDITIONS
TCP CONNECTION STATUS
No violating incidents
Establish TCP connections meeting
requirements to the SoS
Occurrence of 1 st violating
Abort TCP connections having a level of
incident during current
precedence LP = LP1 for a pre-determined
communication session
time interval T1 or a pre-determined number
M1 of the transmissions
Violation incident occurs
Abort TCP connections having a level of
during the time interval T1
precedence LP ≦ LP2 for a pre-determined
or M1 transmissions
time interval T2 or a pre-determined number
M2 of the transmissions
Violation incident occurs
Abort TCP connections having a level of
during the time interval T2
precedence LP ≦ LP3 for a pre-determined
or M2 transmissions
time interval T3 or a pre-determined number
M3 of the transmissions
After the time interval T2 or M2
transmissions, abort the TCP connections
having a level of precedence LP = LP1
Violation incident occurs
Abort TCP connections having a level of
during the time interval T3
precedence LP ≦ LP2 for a pre-determined
or M3 transmissions
time interval T4 or a pre-determined number
M4 of the transmissions.
14. The method of claim 1 , wherein step (e) further comprises:
monitoring compliance of the TCP connections in real time.
15. A method for controlling traffic flow at a network edge proxy of an Internet Protocol (IP) based communication network, comprising software running on said network edge proxy executing instructions to perform the steps of:
(a) associating a speed of service (SoS) with a level of precedence in a Multilevel Precedence and Preemption (MLLP) scheme;
(b) selectively adjusting settings of Transmission Control Protocol (TCP) connections to provide the SoS requirements for a particular level of precedence, said setting consisting of at least one of
(b1) adjusting settings CWND adj for a congestion window;
(b2) adjusting settings RTO adj for a re-transmission time-out time interval using the equation:
RTO adj =min{ N×RTT e ), RTO max },
where N is a pre-determined constant in a range from 1 to 3, RTT e is an estimated transmission round trip time, and RTO max is a maximum re-transmission time-out interval and
where RTO max is determined using the equation:
∑
i
=
1
I
P
e
i
-
1
(
1
-
p
e
)
i
·
[
D
e
+
α
j
e
+
(
i
-
1
)
RTO
max
]
=
L
q
,
where p e is an estimated transmission loss ratio, D e is an estimated transmission transfer delay, j e is an estimated jitter delay, α is a pre-determined constant, I is a number of re-transmissions, i is an integer, 1≦i≦I and L q is a transmission latency requirement; and
(b3) adjusting settings DL adj for time delays for sending acknowledgements of the transmissions;
(c) establishing TCP connections for transmissions meeting the SoS requirements in accordance with levels of precedence of the transmissions, said transmissions being encrypted using High Assurance Internet Protocol Encryption (HAIPE) or an IP Security (IPSec) protocol;
(d) monitoring compliance of the established TCP connections with the MLPP scheme;
(e) temporarily aborting TCP transmissions failing to meet the SoS requirements; and
(f) repeating steps (b)-(f) until the transmission is complete.
16. The method of claim 15 , wherein the network is a wireless, wired, or fiber-optic communication network.
17. The method of claim 15 , wherein step (b1) comprises:
determining the settings CWND adj using an equation:
CWND adj =max{ CWND current ,CWND required },
where CWND current is a current setting for the congestion window, CWND required =RTT e ×MSS)/L q , RTT e is an estimated transmission round trip time, MSS is a maximum size of a transmission unit, and L q is a transmission latency requirement.
18. The method of claim 15 , wherein RTO max is determined using an equation:
RTO max =( L q −D e −αj e )/( I+ 1),
where L q is a transmission latency requirement, D e is a transmission transfer delay, α is a pre-determined constant, j e is a jitter delay, and I is a number re-transmissions.
19. The method of claim 15 , wherein step (b3) comprises:
adjusting the settings DL adj as follows, where RTT e is an estimated transmission round trip time and T m is a pre-determined time interval:
TRAFFIC
CONDITIONS
TIMING FOR SENDING ACKNOWLEDGEMENT
RTT e ≦ T m
Sending acknowledgement with time delay DL adj = 0
RTT e > T m
Sending acknowledgement with time delay
DL adj = RTT e − T m .
20. The method of claim 15 , wherein step (c) comprises:
establishing the TCP connections meeting both of the following conditions:
RTO max >βRTT e , and
RWND>CWND required ,
where RTO max is a maximum re-transmission time-out interval; β is pre-determined constant, RTT e is an estimated transmission rout trip time, RWND is a maximum size of a congestion window at a destination of the transmission, and CWND required =(RTT e ×MSS)/L q , RTT e is an estimated transmission round trip time, MSS is a maximum size of a transmission unit, and L q is a transmission latency requirement.
21. The method of claim 20 , wherein RTO max is determined using an equation:
∑
i
=
1
I
P
e
i
-
1
(
1
-
P
e
)
i
·
[
D
e
+
α
J
e
+
(
i
-
1
)
RTO
max
]
=
L
q
,
where P e is an estimated transmission loss ratio, D e is an estimated transmission transfer delay, J e is an estimated jitter delay, α is a pre-determined constant, I is a number of re-transmissions, i is an integer, 1≦i≦I, and L q is a transmission latency requirement.
22. The method of claim 20 , wherein RTO max is determined using an equation:
RTO max =L a −D e −αJ e )/( I+ 1),
where L a is a transmission latency requirement, D e is a transmission transfer delay, α is a pre-determined constant, J e is a jitter delay, and I is a number re-transmissions.
23. The method of claim 15 , wherein step (e) further comprises:
dynamically controlling the TCP connections as follows, wherein Level of Precedence 1 (LP1)<Level of Precedence 2 (LP2)<Level of Precedence 3 (LP3):
TRAFFIC CONDITIONS
TCP CONNECTION STATUS
No violating incidents
Establish TCP connections meeting
requirements to the SoS
Occurrence of 1 st violating
Abort TCP connections having a level of
incident during current
precedence LP = LP1 for a pre-determined
communication session
time interval T1 or a pre-determined number
M1 of the transmissions
Violation incident occurs
Abort TCP connections having a level of
during the time interval T1
precedence LP ≦ LP2 for a pre-determined
or M1 transmissions
time interval T2 or a pre-determined number
M2 of the transmissions
Violation incident occurs
Abort TCP connections having a level of
during the time interval T2
precedence LP ≦ LP3 for a pre-determined
or M2 transmissions
time interval T3 or a pre-determined number
M3 of the transmissions
After the time interval T2 or M2 trans-
missions, abort the TCP connections having a
level of precedence LP = LP1
Violation incident occurs
Abort TCP connections having a level of
during the time interval T3
precedence LP ≦ LP2 for a pre-determined
or M3 transmissions
time interval T4 or a pre-determined number
M4 of the transmissions.
24. The method of claim 15 , wherein step (d) further comprises:
monitoring compliance of the TCP connections in a real time.