Apparatus for monitoring state of traffic in telecommunication network

• US 6,147,974 A
• Filed: 02/27/1998
• Issued: 11/14/2000
• Est. Priority Date: 03/04/1997
• Status: Expired due to Fees

• Current Assignee: KDDI Corporation
• Original Assignee: Kokusai Denshi Denwa Company Limited (KDDI Corporation)
• Inventors: Matsumoto, Kazunori, Hashimoto, Kazuo
• Primary Examiner(s): VU, HUY DUY
• Application Number:
  US09/031,913
• Time in Patent Office: 991 Days
• Field of Search: 370/229, 370/230, 370/235, 370/236, 370/237, 370/241, 370/242, 370/244, 370/250, 370/252, 370/253, 370/254, 379/1, 379/9, 379/10, 379/13, 379/14, 379/15, 379/32, 379/34, 379/133, 379/134, 379/139
• US Class Current: 370/252
• CPC Class Codes: H04M 3/36 Statistical metering, e.g. recording occasions when traffic exceeds capacity of trunks

Abstract

An apparatus for monitoring state of traffic in a telecommunication network by calculating ABR based upon BID and ANS per unit time in the network. This apparatus includes a model generation unit for receiving sampled data for generating models, the sampled data being constituted by a set of pairs of BID and ANS per unit time when the network is normal, and for calculating an average connection ratio when the network is normal based upon the sampled data, a calculation unit for calculating log-likelihood at a single point under hypothesizes of normality and degradation of the network based upon time series BID and ANS and upon the average connection ratio calculated by the model generation unit, and a detection unit for detecting whether state of traffic in the network is normal or degraded based upon time series log-likelihood of under hypothesizes of normality and degradation calculated by the calculation unit.

Claims

1. An apparatus for monitoring state of traffic in a telecommunication network by calculating ABR based upon BID and ANS per unit time in the network, said apparatus comprising:

a model generation means for receiving sampled data for generating models, said sampled data being constituted by a set of pairs of BID and ANS per unit time when the network is normal, and for calculating an average connection ratio when the network is normal based upon the sampled data;

a calculation means for calculating log-likelihood at a single point under hypothesizes of normality and degradation of the network based upon time series BID and ANS and upon the average connection ratio calculated by said model generation means; and

a detection means for detecting whether state of traffic in the network is normal or degraded based upon time series log-likelihood of under hypothesizes of normality and degradation calculated by said calculation means.

2. The apparatus as claimed in claim 1, wherein said model generation means includes means for calculating an average connection ratio when the network is normal p by from an equation of:

space="preserve" listing-type="equation">p=Σ

.sub.i.sup.n ai/Σ

.sub.i.sup.n bi

where a_i denotes time series ANS and b_i denotes time series BID.

3. The apparatus as claimed in claim 1, wherein said calculation means includes means for calculating log-likelihood at a single point under hypothesis of normality from ll(b,a|p) when log-likelihood at a single point ll(b,a|p) is defined as:

##EQU4## where b and a denote measured values of BID and ANS, and p denotes a probability of successful connection.

4. The apparatus as claimed in claim 1, wherein said calculation means includes means for calculating a probability of successful connection when the network is degraded p_W from an equation of:

##EQU5## where b and a denote measured values of BID and ANS, p denotes an average connection ratio when the network is normal, and ε

(0<

ε

<

1) denotes a sensitivity parameter.

5. The apparatus as claimed in claim 1, wherein said model generation means includes means for generating models for each combination of days groups based upon AIC.