Method for monitoring and meeting customer bandwidth demand in operational IP data networks

• US 6,751,664 B1
• Filed: 07/05/2000
• Issued: 06/15/2004
• Est. Priority Date: 07/05/2000
• Status: Expired due to Fees

• Current Assignee: AT&T Corporation (AT&T, Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Kogan, Yaakov, Ramachandran, Gomathi, Maguluri, Gangaji
• Primary Examiner(s): VU, VIET DUY
• Application Number:
  US09/609,215
• Time in Patent Office: 1,441 Days
• Field of Search: 709/217, 709/219, 709/224, 709/225, 709/226, 709/238, 709/239, 709/249, 709/250, 370/229, 370/230, 370/232, 370/234
• US Class Current: 709/224
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 41/5067 Customer-centric QoS measurements
• CPC Class Codes: H04L 43/0829 Packet loss
• CPC Class Codes: H04L 43/0882 Utilisation of link capacity
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 47/12 Avoiding congestion; Recovering from congestion
• CPC Class Codes: H04L 47/15 in relation to multipoint traffic arrangements for broadcast or multicast in data networks H04L12/18
• CPC Class Codes: H04L 47/70 Admission control; Resource allocation
• CPC Class Codes: H04L 47/822 Collecting or measuring resource availability data
• CPC Class Codes: H04L 47/823 Prediction of resource usage
• CPC Class Codes: H04L 47/826 Involving periods of time
• CPC Class Codes: H04L 47/83 based on usage prediction

Abstract

Monitoring of customer bandwidth demand is accomplished by comparing mean access router uplink utilization to a prescribed threshold value that is computed based upon granular measurements of uplink utilization. This provides an indication of whether customer demand is met with a statistical level of confidence. Further, the actually measured access router uplink utilization can be used to make a predictive assessment of whether to provision a new customer on the access router without adversely affecting existing customers.

Claims

1. A method for monitoring an IP data network within which there is at least one access router connected by at least one uplink to another backbone element, comprising the steps of:

measuring, on a granular basis, the bandwidth utilization of each uplink which connects the access router to said other element; and

verifying bandwidth fulfillment at each access router by comparing the measured mean uplink utilization to a prescribed threshold value,

wherein the bandwidth fulfillment is determined in accordance with inequality m+qs<

1, with m and s² being empirical values for mean and variance, respectively, calculated from the measurements of the uplink utilization during peak time intervals, and with q=q_α

, where q_α

is defined by Φ

(q_α

)=1−

α

where Φ

(x) is the standard normal distribution with mean=0 and variance=1, and α

is a parameter between 0 and 1 defined by a desired confidence level (1−

α

)×

100%.

2. The method according to claim 1 wherein the prescribed threshold value for the mean uplink bandwidth utilization is given by 1−

qs.

3. A method for monitoring an IP data network within which there is at least one access router connected by at least one uplink to another backbone element, comprising the steps of:

measuring, on a granular basis, the bandwidth utilization of each uplink which connects the access router to said other element; and

verifying bandwidth fulfillment at each access router by comparing the measured mean uplink utilization to a prescribed threshold value, wherein if the fulfillment remains below the prescribed threshold value, then performing the step of;

determining if a new customer can be provisioned on the access router by establishing a predicted mean uplink utilization in accordance with mean bandwidth usage of existing customers and required bandwidth of the new customer and comparing whether the predicted mean uplink utilization remains below a threshold value,

wherein the predicted uplink utilization (m_pr)is established in accordance with the relationship;

$m_{\mathrm{pr}}=m_x+\frac{m_x}{n}$

wherein n is given by the relationship;

$n=\frac{\mathrm{currently}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{provisioned}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{access}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{bandwidth}}{\mathrm{bandwidth}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{of}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{new}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{customer}}$

and m_x is obtained from the granular bandwidth measurements in the peak interval by fitting a regression model to predict the m_x for x months ahead.

4. A method for monitoring an IP data network within which there is at least one access router connected by at least one uplink to another backbone element, comprising the steps of:

measuring, on a granular basis, the bandwidth utilization of each uplink which connects the access router to said other element; and

verifying bandwidth fulfillment at each access router by comparing the measured mean uplink utilization to a prescribed threshold value, wherein if the fulfillment remains below the prescribed threshold value, then performing the step of;

determining if a new customer can be provisioned on the access router by establishing a predicted mean uplink utilization in accordance with mean bandwidth usage of existing customers and required bandwidth of the new customer and comparing whether the predicted mean uplink utilization remains below a threshold value,

where the actual decision to provision the customer is made by verifying the inequality $m_{\mathrm{pr}}+\mathrm{qs}<br><br>\sqrt{\frac{m_{\mathrm{pr}}}{m}}<<br><br>1$

and wherein if the inequality is satisfied then the new customer is provisioned, but if the inequality is violated then all provisioning on that router should be closed.

5. A method for determining whether to provision a new customer on an access router coupled by at least one uplink to a backbone and accessed by existing customers via at least one downlink, comprising the steps of:

measuring the utilization of each uplink, which connects the access router to the backbone;

verifying bandwidth fulfillment at each access router in accordance with uplink utilization, and if mean uplink utilization remains below a prescribed threshold value, then

determining if a new customer can be provisioned by establishing predicted mean uplink utilization in accordance with mean bandwidth usage of existing customers and required bandwidth of the new customer and comparing whether the predicted mean uplink utilization remains below a prescribed threshold value,

wherein the bandwidth fulfillment is determined in accordance with inequality m+qs<

1, with m and s² being empirical values for mean and variance, respectively, calculated from the measurements of the uplink utilization during peak time intervals, and with q=q_α

, where q_α

is defined by Φ

(q_α

)=1−

α

where Φ

(x) is the standard normal distribution with mean=0 and variance=1, and α

is a parameter between 0 and 1 defined by a desired confidence level (1−

α

)×

100%.

6. The method according to claim 5 wherein the prescribed threshold value for the mean uplink utilization is given by 1−

qs.

7. A method for determining whether to provision a new customer on an access router coupled by at least one uplink to a backbone and accessed by existing customers via at least one downlink, comprising the steps of:

measuring the utilization of each uplink, which connects the access router to the backbone;

verifying bandwidth fulfillment at each access router in accordance with uplink utilization, and if mean uplink utilization remains below a prescribed threshold value, then

determining if a new customer can be provisioned by establishing predicted mean uplink utilization in accordance with mean bandwidth usage of existing customers and required bandwidth of the new customer and comparing whether the predicted mean uplink utilization remains below a prescribed threshold value,

wherein the predicted mean uplink utilization (m_pr)is established in accordance with the relationship;

$m_{\mathrm{pr}}=m_x+\frac{m_x}{n}$

wherein n is given by the relationship;

$n=\frac{\mathrm{currently}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{provisioned}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{access}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{bandwidth}}{\mathrm{bandwidth}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{of}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{new}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{customer}}$

and m_x is obtained from the granular bandwidth measurements in the peak interval by fitting a regression model to predict the m_x for x months ahead.

8. A method for determining whether to provision a new customer on an access router coupled by at least one uplink to a backbone and accessed by existing customers via at least one downlink, comprising the steps of:

measuring the utilization of each uplink, which connects the access router to the backbone;

verifying bandwidth fulfillment at each access router in accordance with uplink utilization, and if mean uplink utilization remains below a prescribed threshold value, then

determining if a new customer can be provisioned by establishing predicted mean uplink utilization in accordance with mean bandwidth usage of existing customers and required bandwidth of the new customer and comparing whether the predicted mean uplink utilization remains below a prescribed threshold value,

where the actual decision to provision the customer is made by verifying the inequality $m_{\mathrm{pr}}+\mathrm{qs}<br><br>\sqrt{\frac{m_{\mathrm{pr}}}{m}}<<br><br>1$

and wherein if the inequality is satisfied then the new customer is provisioned, but if the inequality is violated then all provisioning on that router should be closed.