JP3828825B2 - Client server system and transmission band management device - Google Patents

Description

【００４９】
【数２】

（▲４▼割当可能帯域の検証）
アドミッション制御装置は、ＡＰ_（Ｎ＋１）〜ＡＰ_Ｍが経路Ｂで使用していた帯域を検知する。
アドミッション制御装置は、経路ＢでＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］に使用されている帯域ｘ_l（ｔ）が、
▲１▼再配分後の最終帯域ｚ_{l_f}以下の場合、すなわち、
【００５０】
【数３】

の場合、ＡＰ_ｌの時刻ｔ₀の再割当帯域ｚ_l（ｔ₀）をｚ_{l_f}とする。
【００５１】
他方、アドミッション制御装置は、経路ＢでＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］に使用されている帯域ｘ_l（ｔ）が、
▲２▼再配分後の最終帯域ｚ_{l_f}より大きく、再配分前の事前割当帯域ｚ_{l_i}以下の場合、すなわち、
【００５２】
【数４】

の場合、ＡＰ_ｌの時刻ｔ₀の再割当帯域ｚ_l（ｔ₀）をその時点での割当帯域、すなわち、ｘ_l（ｔ₀）とする。
【００５３】
アドミッション制御装置は、上記のようにして、ＡＰ_（Ｎ＋１）〜ＡＰ_Ｍに対しての再割当を行った後、経路Ｂ上でこれらＡＰ_（Ｎ＋１）〜ＡＰ_Ｍが使用していない帯域を検証する。
【００５４】
ここで▲２▼より、ＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］が最終割当帯域を超過して使用している伝送帯域は、
【００５５】
【数５】

である。
【００５６】
また、ＡＰ_１〜ＡＰ_Ｎに対して最終割当帯域の合計は、
【００５７】
【数６】

となる。
【００５８】
したがって、アドミッション制御装置は、ＡＰ_１〜ＡＰ_Ｎの時刻ｔ₀における割当帯域ｚ_k（ｔ₀）を
【００５９】
【数７】

とする。
【００６０】
そして、アドミッション制御装置は、ＡＰ_ｌ（Ｎ＋１≦ｌ≦Ｍ）の再割当帯域を、サービス終了時（例えば、ＴＶ電話であれば、回線断後）に縮小する。
【００６１】
アドミッション制御装置は、事前に経路Ｂ上に伝送帯域を割り当てられていたＡＰ_ｌ（Ｎ＋１≦ｌ≦Ｍ）が障害発生前に使用していた伝送帯域を維持するように再割当を行う。したがって、本実施の形態によれば、経路Ｂで提供中のサービス（アプリケーション）の品質が保証される。
【００６２】
（▲５▼帯域割当の継続）
アドミッション制御装置は、上記▲４▼の手順を、定期または不定期の時間ごとに繰り返す。ここでは、説明の簡略化のため、一定時間間隔Δｔ毎に繰り返すものとする。したがって、アドミッション制御装置は、以下のような動作をする。
アドミッション制御装置は、時刻ｔ₀＋ｎΔｔ（ｎは正の整数）において、経路ＢでＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］が使用している伝送帯域ｘ_l（ｔ₀＋ＮΔｔ）が
▲１▼再配分後の最終帯域ｚ_{1_f}以下の場合、すなわち、
【００６３】
【数８】

の場合、ＡＰ_ｌの時刻ｔ₀＋ｎΔｔの割当帯域ｚ₁（ｔ₀＋ＮΔｔ）をｚ_{l_f}とする。
【００６４】
他方、アドミッション制御装置は、時刻ｔ₀＋ｎΔｔ（ｎは正の整数）において、経路ＢでＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］が使用している伝送帯域ｘ_l（ｔ₀＋ＮΔｔ）が
▲２▼再配分後の最終帯域ｚ_{1_f}より大きく、再配分前の事前割当帯域ｚ_{1_i}以下の場合、すなわち、
【００６５】
【数９】

の場合、ＡＰ_ｌの時刻ｔ₀＋ｎΔｔの割当帯域ｚ_l（ｔ₀＋ＮΔｔ）をその時点での割当帯域、すなわち、ｚ_l（ｔ₀＋ＮΔｔ）とする。
【００６６】
アドミッション制御装置は、上記のようにして、ＡＰ_（Ｎ＋１）〜ＡＰ_Ｍに対しての再割当を行った後、経路Ｂ上でこれらＡＰ_（Ｎ＋１）〜ＡＰ_Ｍが使用していない帯域を検証する。
【００６７】
ここで▲２▼より、最終割当帯域を超過してＡＰ_ｌ［Ｎ＋１≦ｌ≦Ｍ］が使用している伝送帯域は、
【００６８】
【数１０】

である。
【００６９】
したがって、アドミッション制御装置は、ＡＰ_１〜ＡＰ_Ｎの時刻ｔ₀＋ｎΔｔにおける割当帯域ｚ_k（ｔ₀＋ＮΔｔ）を
【００７０】
【数１１】

とする。
【００７１】
（▲６▼帯域割当手順の終了）
アドミッション制御装置は、各ＡＰ_１〜ＡＰ_Ｎの割当て帯域ｚ_k（ｔ）が最終割当帯域ｚ_{1_f}と同じになった時点で、上記帯域割当作業を終了する。
以上説明したように、本実施の形態によれば、経路Ａで障害が発生した場合には、経路Ｂ上に伝送帯域を割り当てられていたアプリケーションの帯域値を確保した形で、経路Ａで提供されていたアプリケーションに対して帯域が逐次割当てられる。
【００７２】
（実施の形態の詳細な説明）
図５は、本発明に係る実施の形態を詳細に説明する図である。図５に示すように、ＡＰ_１〜ＡＰ_３は、経路Ａでネットワーク障害が発生したため、経路Ａ上で伝送帯域を使用できなくなっている。
【００７３】
各アプリケーションには、ネットワーク障害発生前に、以下の伝送帯域が事前確保帯域として割り当てられていた。
【００７４】
経路Ａ（総伝送帯域はＴＡ＝１００Ｍｂｐｓとする）
ＡＰ_１（ＴＶ電話サービスＡ １呼あたり５Ｍｂｐｓ） ｚ_{1_i}＝５０Ｍｂｐｓ（１０呼分接続可能）
ＡＰ_２（ＴＶ電話サービスＢ １呼あたり１Ｍｂｐｓ） ｚ_{2_i}＝２５Ｍｂｐｓ（２５呼分接続可能）
ＡＰ_３（ＴＶ電話サービスＣ １呼あたり０．５Ｍｂｐｓ） ｚ_{3_i}＝２５Ｍｂｐｓ（５０呼分接続可能）
【００７５】
経路Ｂ（総伝送帯域Ｔ_B＝１００Ｍｂｐｓとする）
ＡＰ_４（ＶｏＤサービスＡ １ストリームあたり０．５Ｍｂｐｓ） ｚ_{4_i}＝５０Ｍｂｐｓ（１００ストリーム分接続可能）
ＡＰ_５（ＶｏＤサービスＢ １ストリームあたり１Ｍｂｐｓ） ｚ_{5_i}＝５０Ｍｂｐｓ（５０ストリーム分接続可能）
【００７６】
ここで、ネットワーク障害発生直前の時点ｔ₀で、各アプリケーションが使用していた帯域は、
ＡＰ_１ ｘ₁（ｔ₀）＝３０Ｍｂｐｓ（６呼分接続）
ＡＰ_２ ｘ₂（ｔ₀）＝２０Ｍｂｐｓ（２０呼分接続）
ＡＰ_３ ｘ₃（ｔ₀）＝１０Ｍｂｐｓ（２０呼分接続）
ＡＰ_４ ｘ₄（ｔ₀）＝３０Ｍｂｐｓ（６０ストリーム分接続）
ＡＰ_５ ｘ₅（ｔ₀）＝３０Ｍｂｐｓ（３０ストリーム分接続）
である。
【００７７】
本実施の形態は、次のようにして、経路Ａで伝送帯域を専有していたＡＰ_１〜ＡＰ_３に、経路Ｂ上の新たな伝送帯域を割り当てる。
【００７８】
本実施の形態は、まず、ＡＰ_１〜ＡＰ_５に対して、経路ＢでＡＰ_４およびＡＰ_５が確保していた伝送帯域の合計１００Ｍｂｐｓを、初めの事前確保帯域の割合に従い、比例配分して割り当てる。
【００７９】
したがって、ＡＰ_１〜ＡＰ_５に対しては、Ｔ_B ＝１００Ｍｂｐｓが、当初の事前確保帯域の比、５０：２５：２５：５０：５０＝２：１：１：２：２に従って割当てられる。
【００８０】
よって、最終的に、ＡＰ_１〜ＡＰ_５に対して割り当てられる伝送帯域（最終割当帯域）の値は、
ＡＰ_１ ｚ_{1_f}＝２５Ｍｂｐｓ、
ＡＰ_２ ｚ_{2_f}＝１２．５Ｍｂｐｓ、
ＡＰ_３ ｚ_{3_f}＝１２．５Ｍｂｐｓ、
ＡＰ_４ ｚ_{4_f}＝２５Ｍｂｐｓ、
ＡＰ_５ ｚ_{5_f}＝２５Ｍｂｐｓ
となる。
【００８１】
しかしながら、直ちにこの最終割当帯域を実行すれば、あるアプリケーションに対しては、使用可能となる伝送帯域が減少し、使用中の呼やサービスなどが中断されてしまうケースが想定される。そこでつぎのような動作を行うものとする。
【００８２】
ここで、経路Ａで障害発生したときに経路ＢではＡＰ_４がｘ₄（ｔ₀）＝３０Ｍｂｐｓ、ＡＰ_５がｘ₅（ｔ₀）＝３０Ｍｂｐｓの帯域を使用中とする。ＡＰ_４およびＡＰ_５の使用帯域は最終割当帯域を超過しているが、本実施の形態にておいては、この経路Ｂを平常リンクとするＡＰ_４およびＡＰ_５が使用している帯域は確保される。
【００８３】
したがって、ＡＰ_４およびＡＰ_５の割当帯域は、ＡＰ_４がｘ₄（ｔ₀）＝３０Ｍｂｐｓ、ＡＰ_５がｘ₅（ｔ₀）＝３０Ｍｂｐｓになる。
【００８４】
一方、ＡＰ_１〜ＡＰ_３に対しては、合計５０Ｍｂｐｓの帯域が最終的に割当てられるはずだが、この時点ではＡＰ_４とＡＰ_５に合計して６０Ｍｂｐｓ割当てられているため、すぐにＡＰ_１〜ＡＰ_３に割振られる帯域は４０Ｍｂｐｓ（＝１００−６０）となる。
【００８５】
したがって、ネットワーク障害が発生した直後の割当て帯域は、各々、ＡＰ_１がｘ₁（ｔ₀）＝２０Ｍｂｐｓ、ＡＰ_２がｘ₂（ｔ₀）＝１０Ｍｂｐｓ、ＡＰ_３がｘ₃（ｔ₀）＝１０Ｍｂｐｓとなる。
【００８６】
当面アドミッション制御装置は、ＡＰ_４およびＡＰ_５に関しては、最終割当帯域の２５Ｍｂｐｓを参照値として、ＡＰ_１〜ＡＰ_３に関しては２０Ｍｂｐｓ，１０Ｍｂｐｓ，１０Ｍｂｐｓを基準としてアドミッション制御を行うものとする。
【００８７】
次に一定時間が経過の後、時刻ｔ₁でＡＰ_４及びＡＰ_５での使用帯域が
ＡＰ_４ ｘ₄（ｔ₁）＝２５Ｍｂｐｓ（ＶｏＤサービスＡが５０ストリーム分）ＡＰ_５ ｘ₅（ｔ₁）＝２７Ｍｂｐｓ（ＶｏＤサービスＡが２７ストリーム分）であったとする。時刻ｔ₁では、ＡＰ_４およびＡＰ_５については新たなストリームが加わることなく、各サービス終了と同時に使用帯域は減少していくはずである。
【００８８】
そこで、アドミッション制御装置は、ＡＰ_１〜ＡＰ_３に対して、新たに発生した帯域８Ｍｂｐｓを比例配分して割当てる。その結果ＡＰ_１〜ＡＰ_３の割当帯域は、
ＡＰ_１ ｚ₁（ｔ₁）＝２４Ｍｂｐｓ、
ＡＰ_２ ｚ₂（ｔ₁）＝１２Ｍｂｐｓ、
ＡＰ_３ ｚ₃（ｔ₁）＝１２Ｍｂｐｓ、
【００８９】
更に一定時間が経過の後、時刻ｔ₂でＡＰ_５での使用帯域がｘ₅（ｔ₂）＝２５Ｍｂｐｓ（ＶｏＤサービスＢが２５ストリーム分）であったとする。時刻ｔ₂でＡＰ_１〜ＡＰ_３に対しての割当帯域は最終割当帯域と一致する。すなわち、
ＡＰ_１ ｚ₁（ｔ₂）＝２５Ｍｂｐｓ（＝ｚ_{1_f}）
ＡＰ_２ ｚ₂（ｔ₂）＝１２．５Ｍｂｐｓ（＝ｚ_{2_f}）
ＡＰ_３ ｚ₃（ｔ₂）＝１２．５Ｍｂｐｓ（＝ｚ_{3_f}）
である。
【００９０】
またＡＰ_４、ＡＰ_５に対しての割当帯域は、
ＡＰ_４ ｚ₄（ｔ₂）＝２５Ｍｂｐｓ（＝ｚ_{4_f}）
ＡＰ_５ ｚ₅（ｔ₂）＝２５Ｍｂｐｓ（＝ｚ_{5_f}）
となる。
【００９１】
したがって、アドミッション制御装置は、時刻ｔ₂で割当帯域変更の手順を終了する。
【００９２】
【発明の効果】
本発明によれば、アプリケーションがネットワークのある経路における伝送帯域を専有するネットワークにおいて、ある経路で障害が発生した場合に、アプリケーションの使用経路を別の経路（迂回リンク）に切り替えた上で、各アプリケーションの使用帯域に応じて、新たに各アプリケーション毎にネットワークの伝送帯域を割り当て、帯域を専有させることができる。
【００９３】
したがって、本発明によれば、障害が発生し、ネットワークの一部の経路が使用不能となった場合であっても、該ネットワーク上でアプリケーションを提供することができる。
【図面の簡単な説明】
【図１】本発明の概略を示す図である。
【図２】図１に記載のネットワーク上で障害が発生した場合の様子を示す図である。
【図３】本発明に係る実施の形態のネットワーク形態を示す図である。
【図４】本発明に係る実施の形態の再割当てを説明する図である。
【図５】本発明に係る実施の形態を詳細に説明する図である。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a client-server system, and in particular, when an application by streaming such as video distribution or two-way video communication is provided on an IP network, each application occupies a transmission band in a certain path on the network. The server system.
[0002]
[Prior art]
Currently, applications provided on networks using IP are mainly best-effort services such as e-mail and Web viewing, but in the future, services that utilize broadband will gradually develop as access networks become faster. It is assumed that
[0003]
If services that utilize broadband are deployed, the frequency of use of applications mainly for streaming, such as videophone and VoD distribution, will increase in the network.
[0004]
When providing an application mainly for streaming in a network, it is desirable for each application to occupy a transmission band in a certain path on the network.
[0005]
Conventionally, for example, by using Diffserv or the like on an IP network, a transmission band in a certain route on the network is dedicated to each application.
[0006]
[Problems to be solved by the invention]
However, in the conventional network, the transmission bandwidth in a certain route on the network can be exclusively used for each application, but this route is used when a part of the route of the network becomes unusable due to an emergency failure. There is a problem that it becomes impossible to provide an application that uses.
[0007]
Therefore, in view of such circumstances, the present invention provides a client-server system capable of providing an application on a network even when a failure occurs and a part of the network becomes unusable. An object is to provide a transmission band management device.
[0011]
[Means for Solving the Problems]
According to the present invention, the above problem is solved by the means described in the claims. That is, according to the first aspect of the present invention, a client in which a detour route to be used when a failure occurs on a transmission band in each route of the network and a network is set in advance for each application provided by the server to the client. In the server system,
[0012]
When the failure occurs, the bypass is applied to the application of the route in which the failure has occurred and the application of the bypass route of the application, based on the transmission band pre-assigned to these applications in each route. A client-server system characterized in that a transmission band in a route is proportionally allocated and a new transmission band (z _{i — f} ) is allocated to each application.
[0013]
The invention according to claim 2 is a client-server system in which for each application provided by a server to a client, a transmission band in each route of the network and a detour route used when a failure occurs on the network are set in advance. A transmission band management device in which
[0014]
When the failure occurs, the bypass is applied to the application of the route in which the failure has occurred and the application of the bypass route of the application, based on the transmission band pre-assigned to these applications in each route. A transmission band management apparatus characterized in that a transmission band in a path is proportionally distributed and a new transmission band (z _{i — f} ) is allocated to each application.
[0015]
According to a third aspect of the present invention, there is provided a client server system in which for each application provided by a server to a client, a transmission band in each route of the network and a detour route used when a failure occurs on the network are set in advance. In
[0016]
When the failure occurs, for the application of the route in which the failure has occurred, a new transmission band in the detour route used by the detour route application from the total transmission band (z) in the detour route The transmission band (z− (Σz _{i — f} ) −Σ (x) obtained by subtracting both the total value (Σz i — _f ) and the total value of the band in use beyond the new transmission band (Σ (x _i −z _{i — f} )) _i −z _{i_f} )),
[0017]
In the path in which the failure has occurred, the proportionally distributed allocation based on the transmission band allocated in advance to the application of the path in which the failure has occurred,
[0018]
On the other hand, in the case of an application that uses a transmission band beyond the new transmission band (z _{i — f} ) for the application of the detour path, the transmission band (x _{i of the} detour path used by the application _). )
[0019]
Assign the new transmission band _(z i_f) for application in use a new transmission band _(z i_f) following transmission band,
[0020]
A total value (Σ (x _i −z _{i — f} )) of a band in use _exceeding a new transmission band is calculated every fixed time, and the above allocation is repeated until this total value becomes zero. It is a client server system.
[0021]
According to a fourth aspect of the present invention, there is provided a client server system in which for each application provided by a server to a client, a transmission band in each path of the network and a detour path used when a failure occurs on the network are set in advance. A transmission band management device in which
[0022]
When the failure occurs, for the application of the route in which the failure has occurred, a new transmission band in the detour route used by the detour route application from the total transmission band (z) in the detour route Transmission band (z− (Σz _{i_f} ) −Σ (x _i ) obtained by subtracting both the total value (Σz _{i_f} ) and the total value of the band in use beyond the new transmission band (Σ (x _i −z _{i_f} )) −z _{i_f} )),
[0023]
In the path in which the failure has occurred, the proportionally distributed allocation based on the transmission band allocated in advance to the application of the path in which the failure has occurred,
[0024]
On the other hand, for the application of the detour path, in the case of an application that is using the transmission band beyond the new transmission band (z _{i_f} ), the transmission band (x _i ) used by the application is assigned. ,
[0025]
Assign the new transmission band _(z i_f) for application in use a new transmission band _(z i_f) following transmission band,
[0026]
A total value (Σ (x _i −z _{i — f} )) of a band in use _exceeding a new transmission band is calculated every fixed time, and the above allocation is repeated until this total value becomes zero. A transmission band management device.
[0027]
According to the present invention, when a failure or the like occurs on the network and the transmission band cannot be exclusively used in a part of the network (for example, between routers), another route (bypass) New transmission band on the path) can be occupied.
[0028]
According to the present invention, when a failure or the like occurs on the network, the value of the transmission band that each application occupies on the network is dynamically changed according to the value of the transmission band used by these applications. Can be reassigned.
[0029]
DETAILED DESCRIPTION OF THE INVENTION
[Outline of the Present Invention]
FIG. 1 is a diagram showing an outline of the present invention.
[0030]
The network described in FIG. 1 is an IP network, and a plurality of servers and clients exist on the IP network.
[0031]
A plurality of paths exist between the server and the client, and each application provided by the server occupies a transmission band used by itself on these paths. Hereinafter, allocating the transmission band on the network to the application is referred to as allocating the transmission band to the application.
[0032]
The admission control device manages the transmission band assigned to each application. Furthermore, the admission control device controls permission / non-permission of use when each application intends to use a transmission band other than the transmission band assigned to itself. This permission / denial is performed based on the bandwidth used assigned to each application.
[0033]
Note that an existing technique (for example, Diffserv) is used for a specific method in which each application occupies a transmission band between each node (router, layer switch, etc.).
[0034]
FIG. 2 is a diagram illustrating a situation when a failure occurs on the network illustrated in FIG.
[0035]
In FIG. 2, a network failure has occurred in the route A on the network. In this case, a new transmission band is allocated to another application, that is, the path B, to the application that has been allocated the transmission band on the path A as follows.
[0036]
(1) The admission control device calculates the final allocated bandwidth of each application by proportionally allocating the total transmission bandwidth value of the path B according to the transmission bandwidth that each application had before the occurrence of the network failure.
[0037]
(2) In the case where an application assigned a transmission band to the route B uses a transmission band that is equal to or greater than the transmission band value assigned after the occurrence of a network failure before the occurrence of the failure. An excess transmission band value is assigned to the application.
[0038]
(3) The admission control device reallocates the transmission band periodically (or irregularly) after a lapse of a certain time after the occurrence of a network failure, and the allocated band of all applications converges to the final allocated band. Until this, the reassignment operation is repeated.
[0039]
[Description of Embodiment]
FIG. 3 is a diagram showing a network configuration of the embodiment according to the present invention, and FIG. 4 is a diagram for explaining reassignment of the embodiment according to the present invention. Hereinafter, the reallocation according to the embodiment of the present invention will be described with reference to FIGS. 3 and 4. In the present embodiment, an example is shown in which the link capacity is relatively free and a detour path for the current application can be secured.
[0040]
In this embodiment, a server client type service (application) is provided in an IP network composed of a plurality of routers and nodes. There are a plurality of paths between the server and the client on the network. Each application has a transmission band assigned to each path.
[0041]
(▲ 1 ▼ Pre-allocated bandwidth setting for each application)
N applications from AP_1 to AP_N (N is a positive integer) are allocated to the path A. Therefore, the route A is a normal link for AP_1 to AP_N.
[0042]
Here, it is _{assumed that} the bandwidth of z k — i is pre-assigned to the k-th application AP_k, and the total of the pre-assigned bandwidths of AP — 1 to AP_N is Z A — _i . The transmission band actually used by the k-th application on the path A at a certain time t is assumed to be x _k (t). The AP_1～AP_N, as a pre-allocated bandwidth, all bands T _A of the path A is allocated. Therefore, T _A = Z _A — _i .
[0043]
On the path B, MN applications from AP_ (N + 1) to AP_M are allocated. That is, for AP_ (N + 1) to AP_M, the path B is a normal link.
[0044]
Here, it is _{assumed that} a band of z _{l_i} is assigned in advance to the l-th application AP_l, and the total of the pre-assigned bands of AP_ (N + 1) to AP_M is Z _{b_i} . The transmission band actually used on the path B by the k-th application at a certain time t is assumed to be x _l (t). The AP_ (N + 1) ~AP_M, as a pre-allocated bandwidth, is allocated all bands T _B of path B. Therefore, T _B = Z _B — _i .
[0045]
(▲ 2 ▼ network failure)
Assume that a network failure occurs in the path A, and communication between the server and the client regarding the AP_1 to AP_N assigned the transmission band on the path A becomes impossible.
[0046]
In this case, the transmission band is reassigned on the path B to AP_1 to AP_N. Therefore, the route B becomes a detour link for AP_1 to AP_N. The reallocation is performed by the admission control device as follows.
[0047]
(3) Calculation of final allocated bandwidth for each application
The admission control device finally distributes the bandwidth Z _{B_i of the} path B to the applications AP_1 to AP_M based on the bandwidth that the AP_1 to AP_M occupied in the path A or the path B before the failure occurred. The allocated bandwidth (final allocated bandwidth) is calculated.
[0048]
[Expression 1]

[0049]
[Expression 2]

(4) Verification of allocatable bandwidth
The admission control device detects the bandwidth used by AP_ (N + 1) to AP_M on path B.
The admission control device uses the band x _l (t) used for AP_l [N + 1 ≦ l ≦ M] in the path B,
(1) If the final band after redistribution is less than or equal to z _{l_f} , that is,
[0050]
[Equation 3]

In this case, the reallocation band z _l (t ₀ ) at time t ₀ of _{AP_l} is set to z _{l_f} .
[0051]
On the other hand, the admission control device has a bandwidth x _l (t) used for AP_l [N + 1 ≦ l ≦ M] in the path B,
(2) When the bandwidth is larger than the final bandwidth z _{l_f} after the reallocation and is equal to or _smaller than the pre-allocation bandwidth z _{l_i} before the reallocation, that is,
[0052]
[Expression 4]

In this case, the reallocation band z _l (t ₀ ) at time t ₀ of AP_l is the allocation band at that time, that is, x _l (t ₀ ).
[0053]
The admission control device verifies the bands that are not used by AP_ (N + 1) to AP_M on the route B after reassigning AP_ (N + 1) to AP_M as described above.
[0054]
Here, from (2), the transmission band that AP_l [N + 1 ≦ l ≦ M] uses exceeding the final allocated band is:
[0055]
[Equation 5]

It is.
[0056]
The total final allocated bandwidth for AP_1 to AP_N is
[0057]
[Formula 6]

It becomes.
[0058]
Therefore, the admission control apparatus uses the allocated bandwidth z _k (t ₀ ) at time t ₀ of AP_1 to AP_N as follows.
[Expression 7]

And
[0060]
Then, the admission control apparatus reduces the reassigned bandwidth of AP_l (N + 1 ≦ l ≦ M) at the end of the service (for example, after a line disconnection in the case of a videophone).
[0061]
The admission control device performs reassignment so that AP_l (N + 1 ≦ l ≦ M), which has been assigned a transmission band on the path B in advance, maintains the transmission band used before the failure occurred. Therefore, according to the present embodiment, the quality of the service (application) being provided on the route B is guaranteed.
[0062]
((5) Continuation of bandwidth allocation)
The admission control device repeats the procedure (4) above at regular or irregular intervals. Here, for simplification of description, it is assumed to repeat every certain time interval Δt. Therefore, the admission control device operates as follows.
The admission control device reallocates the transmission band x _l (t ₀ + NΔt) used by AP_l [N + 1 ≦ l ≦ M] in the path B at time t ₀ + nΔt (n is a positive integer). If it is less than or equal to the last final band z _{1_f} , that is,
[0063]
[Equation 8]

In this case, the allocation band z ₁ (t ₀ + NΔt) of AP_l at time t ₀ + nΔt is defined as z _{l_f} .
[0064]
On the other hand, in the admission control device, at time t ₀ + nΔt (n is a positive integer), the transmission band x _l (t ₀ + NΔt) used by AP_l [N + 1 ≦ l ≦ M] in the path B is (2). When the bandwidth is larger than the final bandwidth z _{1_f} after the reallocation and is equal to or less than the pre-allocation bandwidth z _{1_i} before the reallocation,
[0065]
[Equation 9]

In this case, the allocation band z _l (t ₀ + NΔt) of AP_l at time t ₀ + nΔt is the allocation band at that time, that is, z _l (t ₀ + NΔt).
[0066]
The admission control device verifies the bands that are not used by AP_ (N + 1) to AP_M on the route B after reassigning AP_ (N + 1) to AP_M as described above.
[0067]
Here, from (2), the transmission band used by AP_l [N + 1 ≦ l ≦ M] exceeding the final allocated bandwidth is
[0068]
[Expression 10]

It is.
[0069]
Therefore, the admission control apparatus uses the allocated bandwidth z _k (t ₀ + NΔt) at time t ₀ + nΔt of AP — 1 to AP_N.
[Expression 11]

And
[0071]
(6) End of bandwidth allocation procedure
The admission control device ends the bandwidth allocation work when the allocated bandwidth z _k (t) of each AP_1 to AP_N becomes the same as the final allocated bandwidth z _{1_f} .
As described above, according to the present embodiment, when a failure occurs in the route A, the bandwidth value of the application assigned the transmission band on the route B is secured and provided in the route A. Bandwidth is sequentially allocated to the applications that have been made.
[0072]
(Detailed description of embodiment)
FIG. 5 is a diagram illustrating the embodiment according to the present invention in detail. As shown in FIG. 5, AP_1 to AP_3 cannot use the transmission band on the route A because a network failure has occurred on the route A.
[0073]
Each application was assigned the following transmission band as a pre-reserved band before the occurrence of a network failure.
[0074]
Route A (Total transmission bandwidth is TA = 100 Mbps)
AP_1 (TV phone service A 5 Mbps per call) z _{1_i} = 50 Mbps (connectable for 10 calls)
AP_2 (TV phone service B 1 Mbps per call) z _{2_i} = 25 Mbps (25 calls can be connected)
AP_3 (TV phone service C 0.5 Mbps per call) z _{3_i} = 25 Mbps ( 50 calls can be connected)
[0075]
Path B (total transmission band T _B = 100 Mbps)
AP — 4 (VoD service A 0.5 Mbps per stream) z ₄ — _i = 50 Mbps (100 streams connectable)
AP — 5 (VoD service B 1 Mbps per stream) z ₅ — _i = 50 Mbps (50 streams connectable)
[0076]
Here, the bandwidth used by each application at time t ₀ immediately before the occurrence of the network failure is
AP_1 x ₁ (t ₀ ) = 30 Mbps (connection for 6 calls)
AP_2 x ₂ (t ₀ ) = 20 Mbps (connection for 20 calls)
AP — ₃ × ₃ (t ₀ ) = 10 Mbps (20 call connection)
AP — ₄ × ₄ (t ₀ ) = 30 Mbps (60 streams connected)
AP — ₅ × ₅ (t ₀ ) = 30 Mbps (connection for 30 streams)
It is.
[0077]
In the present embodiment, a new transmission band on path B is allocated to AP_1 to AP_3 that exclusively occupied the transmission band on path A as follows.
[0078]
In the present embodiment, first, a total of 100 Mbps of transmission bands reserved by AP_4 and AP_5 in the path B is allocated to AP_1 to AP_5 in proportion to the initial pre-reserved band.
[0079]
Accordingly, for AP_1 to AP_5, T _B = 100 Mbps is allocated according to the ratio of the initial pre-reserved bandwidth, 50: 25: 25: 50: 50 = 2: 1: 1: 2: 2.
[0080]
Therefore, finally, the value of the transmission band (final allocation band) allocated to AP_1 to AP_5 is
AP_1 z _{1_f} = 25 Mbps,
AP — 2 z 2 — _f = 12.5 Mbps,
AP — 3 z 3 — _f = 12.5 Mbps,
_{AP_} 4 z 4_f = 25Mbps,
AP — 5 z 5 — _f = 25 Mbps
It becomes.
[0081]
However, if this final allocated bandwidth is executed immediately, the transmission bandwidth that can be used for a certain application decreases, and a call or service that is in use may be interrupted. Therefore, the following operation is performed.
[0082]
Here, it is assumed that, when a failure occurs in the path A, the path B is in use with a band of AP_4 x ₄ (t ₀ ) = 30 Mbps and AP_5 x ₅ (t ₀ ) = 30 Mbps. The band used by AP_4 and AP_5 exceeds the final allocated band. However, in this embodiment, the band used by AP_4 and AP_5 having the path B as a normal link is secured.
[0083]
Therefore, the allocated bandwidths of AP_4 and AP_5 are such that AP_4 is x ₄ (t ₀ ) = 30 Mbps and AP_5 is x ₅ (t ₀ ) = 30 Mbps.
[0084]
On the other hand, a total bandwidth of 50 Mbps should be finally assigned to AP_1 to AP_3, but at this point, AP_4 and AP_5 are assigned a total of 60 Mbps, so the bandwidth immediately allocated to AP_1 to AP_3 is 40 Mbps (= 100-60).
[0085]
Therefore, the allocated bandwidth immediately after the occurrence of the network failure is AP ₁ for x ₁ (t ₀ ) = 20 Mbps, AP_2 for x ₂ (t ₀ ) = 10 Mbps, and AP_ ₃ for x ₃ (t ₀ ) = 10 Mbps.
[0086]
For the time being, the admission control apparatus shall perform admission control with reference to AP_4 and AP_5 with reference to the final allocated bandwidth of 25 Mbps and AP_1 to AP_3 with reference to 20 Mbps, 10 Mbps, and 10 Mbps.
[0087]
Next, after a certain period of time has elapsed, at time t ₁ , the bandwidth used by AP_4 and AP_5 is AP_4 x ₄ (t ₁ ) = 25 Mbps (50 streams of VoD service A) AP_5 x ₅ (t ₁ ) = 27 Mbps (VoD Assume that service A is 27 streams). At time t ₁ , for AP_4 and AP_5, a new stream is not added, and the bandwidth to be used should decrease at the end of each service.
[0088]
Therefore, the admission control device allocates a newly generated bandwidth of 8 Mbps in proportion to AP_1 to AP_3. As a result, the allocated bandwidth of AP_1 to AP_3 is
AP_1 z ₁ (t ₁ ) = 24 Mbps,
AP_2 z ₂ (t ₁ ) = 12 Mbps,
AP — ₃ z ₃ (t ₁ ) = 12 Mbps,
[0089]
Further, assume that the bandwidth used at AP_5 is x ₅ (t ₂ ) = 25 Mbps (VoD service B is for 25 streams) at a time t ₂ after a certain time has passed. Allocated bandwidth against AP_1~AP_3 at time t ₂ is consistent with the final allocation band. That is,
AP — ₁ z ₁ (t ₂ ) = 25 Mbps (= z 1 — _f )
AP — ₂ z ₂ (t ₂ ) = 12.5 Mbps (= z ₂ — _f )
AP — ₃ z ₃ (t ₂ ) = 12.5 Mbps (= z ₃ — _f )
It is.
[0090]
The allocated bandwidth for AP_4 and AP_5 is
AP — ₄ z ₄ (t ₂ ) = 25 Mbps (= z ₄ — _f )
AP — ₅ z ₅ (t ₂ ) = 25 Mbps (= z ₅ — _f )
It becomes.
[0091]
Thus, admission control device ends the steps of allocated band change at time t _2.
[0092]
【The invention's effect】
According to the present invention, in a network where an application occupies a transmission band in a certain route of the network, when a failure occurs in a certain route, the application use route is switched to another route (detour link). A network transmission band can be newly assigned to each application according to the use band of the application, and the band can be exclusively used.
[0093]
Therefore, according to the present invention, an application can be provided on a network even when a failure occurs and a part of the route of the network becomes unusable.
[Brief description of the drawings]
FIG. 1 is a diagram showing an outline of the present invention.
FIG. 2 is a diagram illustrating a situation when a failure occurs on the network illustrated in FIG. 1;
FIG. 3 is a diagram showing a network configuration according to the embodiment of the present invention.
FIG. 4 is a diagram for explaining reassignment according to the embodiment of the present invention;
FIG. 5 is a diagram illustrating an embodiment according to the present invention in detail.

Claims (4)

In a client server system in which a detour route to be used when a failure occurs on a transmission band in each route of the network and a network for each application provided to the client by the server,
When the failure occurs,
To the application of the path where the failure has occurred and the application of the detour path of the application,
In each route, based on the transmission band pre-assigned to these applications,
A client-server system, wherein a transmission band in the detour path is proportionally distributed and a new transmission band (z _{i — f} ) is allocated to each application. A transmission band management device in a client server system in which a transmission band in each path of a network and a detour path to be used when a failure occurs on the network are preset for each application provided to a client by a server. ,
When the failure occurs,
To the application of the path where the failure has occurred and the application of the detour path of the application,
In each route, based on the transmission band pre-assigned to these applications,
A transmission band management apparatus characterized by allocating a transmission band in the detour path proportionally and assigning a new transmission band (z _{i — f} ) to each application. In a client server system in which a detour route to be used when a failure occurs on a transmission band in each route of the network and a network for each application provided to the client by the server,
When the failure occurs,
For the application of the route where the failure occurred,
From the total transmission band (z) in the detour path, the total value (Σz _{i_f} ) of the new transmission band in the detour path used by the application of the detour path and the total band in use beyond the new transmission band A transmission band (z− (Σz _{i_f} ) −Σ (x _i −z _{i_f} )) obtained by subtracting both values (Σ (x _i −z _{i_f} )),
In the path in which the failure has occurred, the proportionally distributed allocation based on the transmission band allocated in advance to the application of the path in which the failure has occurred,
On the other hand, in the case of an application that uses a transmission band beyond the new transmission band (z _{i — f} ) for the application of the detour path, the transmission band (x _{i of the} detour path used by the application _). )
Assign the new transmission band _(z i_f) for application in use a new transmission band _(z i_f) following transmission band,
A total value (Σ (x _i −z _{i — f} )) of a band in use _exceeding a new transmission band is calculated every fixed time, and the above allocation is repeated until this total value becomes zero. Client server system. A transmission band management device in a client server system in which a transmission band in each path of a network and a detour path to be used when a failure occurs on the network are preset for each application provided to a client by a server. ,
When the failure occurs,
For the application of the route where the failure occurred,
From the total transmission band (z) in the detour path, the total value (Σz _{i_f} ) of the new transmission band in the detour path used by the application of the detour path and the total band in use beyond the new transmission band value _(Σ (x _i -z i_f) transmission band obtained by subtracting both the _{(z- (Σz i_ f) -Σ} (x i -z i_f)),
In the path in which the failure has occurred, the proportionally distributed allocation based on the transmission band allocated in advance to the application of the path in which the failure has occurred,
On the other hand, in the case of an application that uses a transmission band beyond the new transmission band (z _{i — f} ) for the application of the detour path, the transmission band (x _{i of the} detour path used by the application _). )
Assign the new transmission band _(z i_f) for application in use a new transmission band _(z i_f) following transmission band,
A total value (Σ (x _i −z _{i — f} )) of a band in use _exceeding a new transmission band is calculated every fixed time, and the above allocation is repeated until this total value becomes zero. Transmission band management device.

Images