DISTRIBUTION OF A VIRTUAL NETWORK OPERATOR'S SUBSCRIPTION IN A SATELLITE SPOTBEAM NETWORK

Abstract

A method for bandwidth management for a host network operator (HNO), wherein the method comprises the following: Providing shared beams shared by two or more of a multitude of virtual network operators (VNOs), each VNO having a subscription that includes a global bandwidth limit applicable for a flow control epoch for each VNO, and an end-device subscription for each end-device associated with that VNO; Aggregating demand and endpoint subscriptions of active endpoints per shared beam per VNO in the current flow control epoch; Distributing the aggregated demand and subscriptions of active end devices for each shared beam per VNO into a distribution bandwidth per shared beam per VNO, where the distribution bandwidth is based on the respective limit of the global bandwidth of the respective VNO; Proportional compensation of an overdrawn distribution bandwidth per shared beam per VNO based on proportionalities of limiting the global bandwidth of each VNO subscribing to the shared beam, for each shared beam, to provide a proportional distribution bandwidth per shared beam per VNO; and the subscription also includes a guaranteed global minimum bandwidth, a data flow from the active end devices is specified as either a guaranteed or best-effort data flow, The aggregation further includes aggregating the demand for and subscriptions of active endpoints for data flows that are stated to be guaranteed as one guaranteed demand and one guaranteed subscription of active endpoints for each shared beam per VNO, and The distribution also includes setting the bandwidth allocation for each of the shared beams per VNO to be equal to or greater than the respective guaranteed demand and subscriptions of active terminal devices for each shared beam per VNO.

Description

AREA

[0001] The present teachings disclose a method and system where a virtual network operator (VNO) can subscribe to aggregated bandwidth at the system level for a global subscription across multiple satellites or at a single satellite level. In some embodiments, a VNO can subscribe to bandwidth at the gateway level or even at the beam level. In particular, a system-level subscription across multiple satellites, a system-level subscription for a satellite, a gateway-scope subscription, or a beam-level subscription, among others, can be provided to various VNOs in a network. GENERAL STATE OF THE ART

[0002] The use of a high-throughput satellite (HTS) using spot beam technology has emerged in commercial satellite communications over the past decade. A significant capacity increase has been achieved with an HTS with frequency reuse across multiple, tightly focused spot beams. However, due to the high number of spot beams, an HTS uses geographically distributed gateways or teleports, and multiple HTSs are deployed within a system to cover more geographic regions.

[0003] In a wholesale-reseller network, a virtual network operator (VNO) purchases bandwidth resources from a host network operator (HNO), typically without investing in the physical infrastructure. The state-of-the-art subscription model consists of various parameters statically provisioned for each VNO per beam per gateway. The parameters per beam per gateway are managed by beam bandwidth managers on the gateway to limit the beam usage of each VNO per subscription parameter of that beam and by proportionally partitioning oversubscriptions within the beam. Due to the localized bandwidth management for each beam, some of a VNO's demands may be subject to congestion on one beam while free capacity exists on another beam to which the VNO also subscribes.Therefore, localized bandwidth management cannot dynamically respond to the overall requirements of the VNO and the system capacity to which the VNO has subscribed.

[0004] US 2016 / 072574A1 discloses systems and methods for bandwidth management and allocation between terminal groups (TGs) in a shared access network. The method comprises receiving an allocated bandwidth from the bandwidth manager based on the bandwidth allocation for the multiple TGs and the associated IPGW instances, and allocating this bandwidth to each IPGW instance. The allocation of the total available bandwidth among the multiple TGs is performed in such a way that it proportionally covers the communication needs of each of the multiple TGs.

[0005] US 2016 / 072691A1 discloses a method for bandwidth management, in particular for managing bandwidth allocation for terminal devices. It describes how the terminal devices can be geographically separated from one another while being logically grouped and covered by the beams of different satellites, whereby the broadband network can, in particular, be a satellite network. A throughput limit (or subscribed rate) is specified for the terminal groups (TGs). Furthermore, the allocation of slots is presented as proportionally balanced with respect to the radio resources if the average frequency efficiency is the same for all TGs.

[0006] US 2014 / 244808A1 describes a virtual network operator function (VNO) for the management control of virtual network services. The described satellite communications system comprises satellites that enable the connection between multiple satellite terminals (VSATs), as well as subsystems in the form of a network data plane and a network management plane. Depending on various factors, such as the modulation and coding methods used and the distributions, the portion of the spectrum supported by the gateway server enables a specific bandwidth, which is provided in megabits and can then be distributed among a number of virtual networks. SUMMARY

[0007] This Summary is provided to introduce a selection of concepts in simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter. The scope of the claimed subject matter is defined by independent claims 1 and 10. The dependent claims relate to preferred embodiments of the invention.

[0008] The present teachings disclose a method and system where a virtual network operator (VNO) can subscribe to aggregated system-level bandwidth with, among other things, a system-level subscription across multiple satellites, a system-level subscription on a satellite, a gateway-scope subscription, or a beam-level subscription. The present subscription model provides a separate set of parameter values ​​for the in-route (uplink) and out-route (downlink) directions, including parameter values ​​for a maximum bandwidth, a minimum bandwidth, real-time bandwidth, and multicast bandwidth. In some embodiments, the multicast bandwidth subscription parameter applies only to the out-route. In some embodiments, the real-time bandwidth and multicast bandwidth parameters are only available for a beam-level subscription.

[0009] A method for bandwidth management for a host network operator (HNO) is disclosed. The method includes: providing shared beams shared by two or more of a plurality of virtual network operators (VNOs), each VNO having a subscription including a global bandwidth limit applicable to a flow control epoch for each VNO, and a terminal subscription for each terminal associated with the respective VNO; aggregating demand and terminal subscriptions of active terminals per shared beam per VNO in the current flow control epoch; distributing the aggregated demand and active terminal subscriptions for each shared beam per VNO into a distribution bandwidth per shared beam per VNO, the distribution bandwidth based on the respective global bandwidth limit of the respective VNO;Proportionately balancing any oversubscription of the distribution bandwidths per shared beam per VNO based on the proportionalities of the global bandwidth limitation of each VNO subscribing to the shared beam for each shared beam to provide a proportional distribution bandwidth per shared beam per VNO;and wherein the subscription further comprises a guaranteed global minimum bandwidth, a data flow from the active end devices is specified as either a guaranteed or best-effort data flow, the aggregating further comprises aggregating the demand for and subscriptions of active end devices for data flows specified as being guaranteed as a guaranteed demand and a guaranteed subscription of active end devices for each shared beam per VNO, and the distributing further comprises setting the bandwidth distribution for each of the shared beams per VNO to be equal to or greater than the respective guaranteed demand and subscriptions of active end devices for each shared beam per VNO.;

[0010] A system for performing bandwidth management for a host network operator (HNO) is disclosed. The system includes: shared beams shared by two or more of a plurality of virtual network operators (VNOs), each VNO including a subscription that includes a global bandwidth limit applicable to a flow control epoch for each VNO and an end device subscription for each end device assigned to the respective VNO; a beam-level bandwidth manager for aggregating demand and end device subscriptions of active end devices per shared beam per VNO in the current flow control epoch; and a global bandwidth manager (GBM). The GBM is operateddistribute the aggregated demand and subscriptions of active end devices for each shared beam per VNO into a distribution bandwidth per shared beam per VNO, where the distribution bandwidth is based on the respective global bandwidth limit of the respective VNO; and proportionally balance an oversubscription of the distribution bandwidths per shared beam per VNO based on the proportionalities of the global bandwidth limit of each VNO subscribing to the shared beam for each shared beam, in order to provide a proportional distribution bandwidth per shared beam per VNO, where the subscription includes a guaranteed minimum global bandwidth, a data flow is specified by the active devices as either a guaranteed or best-effort data flow, the beam bandwidth manager is further operated, the demand for and subscriptions of active end devices for data flows,which are stated to be guaranteed, as a guaranteed demand and a guaranteed subscription of terminals for each shared beam per VNO, and the GBM is further operated to set the bandwidth distribution for each of the shared beams per VNO as equal to or greater than the respective guaranteed demand and subscriptions of active terminals for each shared beam per VNO.

[0011] Additional features are set forth in the following description and are partly apparent from the description or can be learned by practicing what is described. DRAWINGS

[0012] To describe how the above and other advantages and features can be achieved, a more detailed description is given below, presented with reference to specific embodiments thereof, which are illustrated in the accompanying drawings. While understanding that these drawings represent only typical embodiments and are therefore not to be considered limiting of the scope of protection, implementations are described and explained with additional specificity and detail by the accompanying drawings. Fig. 1 illustrates an exemplary system for performing bandwidth management for a host network operator (HNO) according to various embodiments. Fig. 2 illustrates communication flows on an outroute chain according to various embodiments. Fig. 3 illustrates communication flows on an in-route chain according to various embodiments. Fig. 4 illustrates a flowchart of an example method for bandwidth management for an ENT according to some embodiments.

[0013] In the drawings and detailed description, the same drawing reference numerals refer to the same elements, features, and structures unless otherwise described. The relative size and representation of these elements may be exaggerated for clarity, illustration, and convenience. DETAILED DESCRIPTION

[0014] Embodiments are discussed in detail below. While specific implementations are discussed, it is understood that this is for illustrative purposes only. One of ordinary skill in the relevant art will recognize that other components and configurations may be utilized without departing from the spirit and scope of the subject matter of this disclosure.

[0015] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present disclosure. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. Furthermore, the use of the terms "a," "an," etc., does not imply a limitation of quantity, but rather implies the presence of at least one of the recited items. The use of the terms "first," "second," and the like does not imply a particular order; rather, they are included either to identify individual elements or to distinguish one element from another.It should be further understood that the terms "comprises" and / or "comprising" or "includes" and / or "including," when used in this specification, indicate the presence of specified features, regions, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more additional features, regions, integers, steps, operations, elements, components, and / or groups thereof. Although some features may be described with respect to individual example embodiments, aspects need not be so limited, such that features of one or more embodiments may be combined with other features of one or more embodiments.

[0016] In a wholesale reseller network, a virtual network operator (VNO) purchases bandwidth resources from a host network operator (HNO). The present teachings disclose a method and system where a virtual network operator (VNO) can subscribe to aggregated system-level bandwidth with a system-level subscription across multiple satellites, a system-level subscription on a single satellite, a gateway-scope subscription, or a beam-level subscription, among others.

[0017] The present teachings disclose a flexible network configuration for interconnecting an access channel to multiple network operators while also interconnecting a network operator to multiple access channels. Such an interconnected configuration can achieve improvements in network spectrum utilization, load balancing, mobility management, and Quality of Service (QoS) provision. However, the interconnected system introduces new challenges for bandwidth management.

[0018] According to various embodiments, a satellite network with multiple spot beams under the coverage of one or more satellites may be configured as an interconnected network. In such a satellite network, an access channel with shared communication from a ground station to a terminal device is referred to as an outroute or outroute stream; conversely, a communication channel from the terminal device to the gateway is referred to as an inroute or inroute stream. A network gateway, for example, an Internet Protocol (IP) gateway (IPGW), may be located at the ground station. The gateway is a processing component that performs traffic management and routing for multiple terminal devices that share a data stream. Terminal devices connected to the interconnected network may be located / distributed across coverage areas of one or more spot beams provided by one of several satellites in the satellite network.A VNO can be served by one or more beams. Each beam can serve one or more VNOs. A gateway can include a beam-level outroute bandwidth manager (OBM). A gateway can include a beam-level inroute bandwidth manager (IBM). A global bandwidth manager (GBM) can be used to perform bandwidth management for VNOs served by the beams and gateways. The GBM can be located at a central location, for example, near a gateway. The GBM platform and / or a GBM application can be redundant for availability.

[0019] The present teachings disclose methods for bandwidth management of a multi-VNO multi-beam system for both the in-route and out-route directions. In particular, there may be X beams and Y VNOs, where X and Y are greater than one (1), thereby forming an XY model. X beams may be located in multiple beams of one or more satellites. Y VNOs may be located at one or more network operators. Such an XY system may be fully or partially interconnected, i.e., all X beams may be operational or not for all Y VNOs, and vice versa. In embodiments, the present teachings enable dynamic configuration of the XY system; in other words, the values ​​of X and Y may be varied not only during system deployment but also during operation of the interconnected network.

[0020] In some embodiments, the VNOs may not share beams or gateways. In some embodiments, the beams may serve different operators using different physical resources, such as gateways, satellites, wireless access channels, or wired access channels for each VNO. However, the present teachings are compatible with these prior art subscription solutions.

[0021] In embodiments, the present teachings disclose a centralized bandwidth distribution process performed by a global bandwidth manager (GBM) to balance oversubscription, overallocation, and dynamic distribution of oversubscription by the VNOs across the beams. In embodiments, the present teachings disclose a centralized bandwidth management and reallocation scheme performed by a GBM to manage overallocation of bandwidth capacity by the HNO to the VNOs' subscriptions and to balance a VNO's subscription-based bandwidth allocation using, for example, proportional distribution across beams. The present teachings provide flexibility for bandwidth allocation with differentiated criteria for mixed applications, such as regular data service and guaranteed bandwidth provisioning.The present teachings are adaptable to different user groups, such as consumers, enterprises, and mobility, by sharing the same network resource. The present interconnected multi-beam multi-VNO system exhibits superior performance. The present multi-beam multi-VNO system improves the utilization and flexibility of network resources compared to a prior art subscription system that only provides static beam-level subscriptions to a VNO. The methods and systems of the present teachings perform bandwidth management in an interconnected multi-beam multi-VNO system with improved network resource utilization, superior stability, and flexibility in configuring network resources.

[0022] In a wholesale reseller network, the VNO purchases bandwidth resources from a host network operator (HNO), where a subscription model consists of various in-route and out-route parameters for a VNO, including maximum bandwidth, minimum bandwidth, real-time bandwidth, and multicast bandwidth. In some embodiments, some parameters may apply only to an in-route or out-route. For example, the multicast bandwidth subscription parameter may apply only to the out-route.

[0023] In some embodiments, the minimum bandwidth of a VNO in the network is available even in the worst-case connection state. In some embodiments, the maximum bandwidth of the HNO network is oversubscribed among VNOs.

[0024] In embodiments, the GBM dynamically distributes a VNO's global subscription among its subscribed beams in a proportionally fair manner using various parameters. These parameters may be, for example, per beam for the number of end devices aggregated by QoS type, aggregated guaranteed subscriptions of different QoS types, aggregated maximum subscriptions of different QoS types, and the like. The GBM may distribute each VNO's global subscription among the VNO's subscribed beams. In embodiments, the GBM performs subscription distribution for each VNO independently of other VNOs. After performing subscription distribution, in some embodiments, the GBM may aggregate the bandwidth distributions it has calculated for each VNO within each beam (including any static VNO subscriptions configured to utilize the respective beam).For a maximum subscription of each VNO, the GBM may calculate an oversubscription ratio for each beam, using, for example, the estimated current beam capacity, and the GBM may balance the oversubscription. In some embodiments, the GBM may balance the aggregate minimum subscription across beams after distributing the guaranteed minimum subscription for each VNO among the VNO's subscribed beams. The GBM may check whether the aggregate minimum subscription across VNOs within a beam exceeds the worst-case capacity and balance or adjust the bandwidth across beams if the subscription exceeds the beam capacity.

[0025] In embodiments, a beam-level bandwidth manager, such as the OBM and the IBM, may perform actual allocation of VNO bandwidth on a per-beam basis. The beam-level bandwidth managers may operate independently of each beam. In some embodiments, the beam-level bandwidth managers may operate independently of a communication direction, i.e., the inroute may be managed independently of a corresponding outroute. The beam-level bandwidth managers consider the GBM's global VNO subscription distribution in beams and use this as the beam-level subscription of VNOs. In embodiments, the beam-level bandwidth managers enforce the VNO beam subscription limit and may partition the beam bandwidth across the VNOs in case of congestion.

[0026] In the present teachings, a VNO may subscribe to aggregated bandwidth at the system level. The system-level subscription may include a global subscription across multiple satellites or a subscription for only one satellite. In some embodiments, a subscription may be provided to a VNO at the gateway level or at the beam level.

[0027] In some embodiments, a VNO's subscription may be limited to one of the above options, not multiple options simultaneously. In some embodiments, in an ENT network, a system-level subscription across multiple satellites, a system-level subscription on a satellite, a gateway-scope subscription, or a beam-level subscription to different VNOs may be deployed, among others. In embodiments, a VNO's real-time and multicast subscriptions may be statically configured at the beam level to provide instant bandwidth satisfaction for these subscriptions.

[0028] A global subscription model at the system or gateway level offers the following advantages: It is very difficult for a VNO to know exactly how much bandwidth is needed on a particular beam because traffic flow is dynamic; and the operator does not need to configure static beam-level subscription information for each beam.

[0029] The characteristics of a multi-satellite subscription may include: A VNO subscriber provides various bandwidth parameters (maximum and minimum) as a set of values ​​across multiple satellites, indicating from which of the beams radiated by the multiple satellites the VNO wishes to provide services to its terminal equipment.

[0030] The characteristics of a satellite-level subscription may include parameters such as: a VNO subscriber providing different bandwidth parameters (maximum and minimum) separately for each satellite spectrum, specifying from which beams radiated by those satellite spectra the VNO wishes to provide services to its terminals, and specifying from which gateways the VNO wishes to serve the beams, or the like.

[0031] The characteristics of a gateway-level subscription may include parameters such as a VNO subscriber provisioning different bandwidth parameters (maximum and minimum) separately for each gateway capacity, specifying from which beams served by those gateways the VNO wishes to provide services to its end devices, or the like.

[0032] The characteristics of a beam-level subscription can include parameters such as a VNO subscriber providing various bandwidth parameters (maximum, minimum, real-time, etc.) separately for each beam over which the VNO wishes to provide services to its end devices, or the like. These values ​​can be statically configured at the time of virtual network creation and can be manually modified by an operator during system operation. Beam-level subscriptions are not dynamic and are limited, as described above.

[0033] A spot beam network, such as a Ka-beam network, may include many beams. In embodiments, an intelligent bandwidth allocator dynamically allocates a VNO's system-level (including multi-satellite) or gateway-level bandwidth subscription among its subscribed beams. The minimum and maximum bandwidths per beam are calculated by the intelligent bandwidth allocator for each epoch, where each epoch may be on the order of minutes. The allocator works during both uncongested and congested periods. In some embodiments, the intelligent bandwidth allocator may consider the subscribed static bandwidth of the VNOs on the same beam, where some VNOs may have a static bandwidth per beam subscription.

[0034] In embodiments, a virtual network includes end devices with different Quality of Service (QoS) types. In the virtual network, some end devices may be best-effort end devices, and their bandwidth is not guaranteed. In the virtual network, some other end devices may require bandwidth to be guaranteed. Some end devices may subscribe to real-time bandwidth for real-time applications that may have high latency and jitter requirements. In some embodiments, at least a portion of a end device's overall subscription includes the end device's real-time subscription.

[0035] In embodiments, the GBM may focus on elements at a VNO level, where elements may vary across subscribed beams. These elements may include a number of active terminals, a number of terminals of different QoS types, each terminal's subscription to different QoS types, a demand for active terminals, and the like.

[0036] Fig. 1 logically illustrates a VNO subscription system according to various embodiments.

[0037] A VNO subscription system 100 may include a satellite 102 and a satellite 104, as well as a gateway 110 and a gateway 120. Each of the gateways 110, 120 may serve beams from both satellites 102, 104. In some embodiments, a gateway 130 may only serve beams from satellite 104. Each of the gateways 110, 120, and 130 may be a distributed gateway. Each of the gateways 110, 120, and 130 may perform bandwidth partitioning within a beam for one or more VNOs. In some embodiments, bandwidth partitioning among VNOs may be performed in a proportionally fair manner, for example, by a global bandwidth manager (GBM) 140. In embodiments, the GBM distributes global subscriptions of VNOs into beams, and then the OBM or IBM allocates actual bandwidth for VNOs on each beam based on the calculated beam subscription and the weights of the VNOs and beam demand.

[0038] Each of gateways 110, 120, and 130 may provide, include, or host a beam-level outroute bandwidth manager (OBM) to provide bandwidth management for communications originating from a gateway or from networks connected to the gateway communicating over a beam to end devices served by the beam. The OBM may collect and forward outroute demand 146 to the GBM 140. The outroute demand 146 may include an aggregate per VNO per beam outroute subscription, including demand and usage across end devices. The outroute demand 146 may include current capacity information per beam. The GBM calculates an outroute allocation 144 including a bandwidth allocation per VNO per beam and forwards the outroute allocation 144 to the OBM to allocate the beam capacity per outroute allocation 144 to the VNO.

[0039] Each of the gateways 110, 120, and 130 may provide, include, or host an in-route bandwidth manager (IBM) to provide bandwidth management for inbound communications from Very Small Aperture Terminals (VSATs) communicating over a beam with the respective gateway of the beam or networks connected to the respective gateway. The IBM may collect the in-route demand 150 and forward it to the GBM 140. The in-route demand 150 may include an aggregate per VNO per beam in-route subscription, including demand and usage via end devices. The in-route demand 150 may include current capacity information per beam. The GBM calculates an in-route allocation 148 including a bandwidth allocation per VNO per beam and forwards the in-route allocation 148 to the IBM to allocate the beam capacity per in-route allocation 148 to the VNO.

[0040] In embodiments, the VSATs may be mobile devices designed to be easily moved from a first geographical area to a second geographical area, where the first and second geographical areas are illuminated by different beams included in the system 100. In some embodiments, the VSATs are fixed and cannot be moved by an end user. In embodiments, the VSATs may be active on an ad hoc basis, such that demand and a subscription associated with the VSAT must be accounted for ad hoc in the bandwidth distribution. In embodiments, data flows associated with active VSATs may be active or inactive on an ad hoc basis, such that demands by the active data flows must be accounted for ad hoc by the bandwidth distribution.In embodiments, some of the VSATs served by the network may be active at the time, for example, during office hours, and the network may be subject to congestion, so that a bandwidth shortage caused by the congestion must be taken into account in the bandwidth distribution.

[0041] The global VNO bandwidth distributor or global bandwidth manager (GBM) 140 may be located at a location, for example, at a central location near one of the gateways 110, 120, 130. In some embodiments, the GBM may be co-hosted with one of the gateways 110, 120, 130. The GBM 140 may distribute available capacity of the beams in the VNO subscription system 100 based on a VNO subscription retrieved from a data store 142, an outroute demand 146, and / or an inroute demand 150. In some embodiments, the GBM 140 may perform bandwidth distribution for one or more VNOs within a beam. In some embodiments, the GBM 140 may perform bandwidth distribution for a VNO within a beam.In some embodiments, the GBM 140 may instruct each of the gateways 110, 120, and 130 to perform bandwidth allocation within a beam for the VNOs sharing the beam according to the bandwidth distribution for the VNOs sharing the beam determined by the GBM.

[0042] The intelligent bandwidth allocator or GBM 140 may distribute bandwidth, for example, using proportional fair distribution based on multidimensional factors to calculate how much bandwidth to allocate to each VNO on different beams based on the VNO's subscription. A VNO's subscription level may be established as a system-level, gateway-level, or beam-level subscription. In some embodiments, the GBM 140 may distribute bandwidth based on a quality of service (QoS) requested for communication and the VNO's subscription. For example, the GBM 140 may allocate bandwidth to a VNO by distinguishing between a guaranteed VNO bandwidth subscription and a non-guaranteed (best effort) subscription. In some embodiments, the GBM 140's bandwidth distribution strives for fairness from a quality of service (QoS) perspective for each of the VNOs.

[0043] In embodiments, the GBM 140 assigns a different meaning to one or more of the bandwidth parameters. For example, the subscription parameters minimum and maximum bandwidth are assigned different meanings for a VNO, and the GBM 140 may calculate the minimum and maximum bandwidth separately per beam using different factors and schemes. Embodiment 1: Minimum bandwidth distribution

[0044] A VNO can subscribe to a minimum bandwidth either at the system level or at the gateway level. This bandwidth can be a guaranteed bandwidth for the VNO. The system 100 is dimensioned so that the minimum bandwidth in the network is normally available to all VNOs. In each flow control epoch, the GBM 140 receives information for subscribing to the aggregated, guaranteed bandwidth of all active end devices separately for each beam and each VNO, for example, from the IBM on the gateway 100 or the OBM on the gateway 120. The duration of a flow control epoch for the inroute and outroute can be identical or different. For example, a flow control epoch can be set to every 100 milliseconds (ms) on the outroute, while a flow control epoch for the inroute can be set to every 360 ms. In some embodiments, the inroute flow control epoch can coincide with a superframe marker for the beam.The GBM 140 can perform proportional fair distribution to partition a VNO's minimum system- or gateway-level subscription into its subscribed beams based on the aggregate guaranteed subscription of its active end devices on each beam.

[0045] For a system-level VNO subscription, the GBM 140 can operate as follows. Let Mi,j be the aggregate guaranteed subscription of all active terminals of a VNO i in a beam j at any time, and Msi be the minimum system-level subscription of a VNO i. Then, the distribution of this VNO's minimum subscription on beam j can be calculated as follows, while enforcing that the summation of the minimum bandwidth distribution in beams for a VNO does not exceed the minimum global system-level subscription of the VNO: Mi,j=(Msi*Mi,j) / (Summation of Mi,j over all rays j=i to J)

[0046] For a gateway-level VNO subscription, the GBM 140 can operate as follows. Let Mi,j be the aggregate guaranteed subscription of all active terminals of a VNO i in a beam j at any time, and Mgi be the minimum subscription of a VNO i at the gateway level. Then, the distribution of this VNO's minimum subscription on beam j can be calculated as follows: Mi,j=(Mgi*Mi,j) / (Summation of Mi,j over all rays j=i to J)

[0047] In some embodiments, the present teachings may enforce that the summation of a minimum bandwidth distribution in beams for a VNO does not exceed the minimum gateway-level subscription of the VNOS.

[0048] An exponential moving average (EMA) of Mi,j can be calculated in each flow control epoch over multiple epochs within one update cycle (a few minutes), and at the time of update, the last EMA value can be used as the minimum subscription for each ray.

[0049] If a VNO has all best-effort end devices within a beam, the above scheme calculates a minimum bandwidth of zero (0) for that team. For such a beam, a configured minimum percentage of the total gateway minimum subscription can be calculated, assuming that traffic can arrive on that beam. These values ​​assigned to such a beam type can be borrowed from other beams, and the minimum values ​​for other beams are adjusted accordingly.

[0050] Once all VNO minimum subscriptions are distributed across allowed beams, the process aggregates the minimum subscriptions of all VNOs, including any static minimum subscriptions for each beam. In embodiments, the aggregated minimum may always be available across VNOs within a beam. A worst-case transmission capacity may be configured in data store 142. If the aggregated minimum subscription across end devices within a beam exceeds the worst-case transmission capacity, the minimums for the other beams are adjusted so that a VNO is not deprived of its allowed minimum subscription. Embodiment 2: Maximum bandwidth distribution

[0051] A VNO may subscribe to a maximum or peak bandwidth at the system or gateway level. In some embodiments, this bandwidth may be treated as the best-effort bandwidth. In each flow control epoch, the GBM receives information on aggregated peak bandwidth subscriptions of active end devices separately for each beam and each VNO, as well as the current traffic demand, for example, the outroute demand 146 or the inroute demand 150. The GBM 140 may execute a proportional fairness algorithm to partition a VNO's gateway-level maximum subscription into its subscribed beams based on the aggregated maximum subscriptions of its active end devices and the current demand of active end devices on each beam.

[0052] In embodiments, if the VNO's subscription is at a system level, the system may operate as follows. Let MAXi,j be the aggregated maximum subscription of all active terminals of a VNO i in a beam j at any time, and Di,j be the aggregated current demand of all active terminals of a VNO i in a beam j. Let MAXsi be the system-level maximum subscription of a VNO i, then the distribution of the maximum subscription of this VNO on beam j can be calculated as follows: MAXi,j=(MAXsi*(alpha*Mi,j+(1−alpha)*Di,j) / (Summation of(alpha*Mi,j+(1−alpha)*Di,j) over all rays j=i to J)

[0053] A static factor alpha can provide a deviation between the configured maximum subscription and actual demand in the above calculation. By default, the alpha value can give more weight to demand; for example, alpha can be set to 0.1, 0.2, 0.3, or the like. The present teachings can enforce that the summation of the maximum bandwidth distribution in beams for a VNO does not exceed its maximum subscription at the global system level based on the EMA of Mi,j and Di,j.

[0054] For a gateway-level VNO subscription, the present teachings can be implemented as follows. Let Maxi,j be the aggregate maximum subscription of all active terminals of a VNO i in a beam j at any time, and Di,j be the aggregate current demand of all active terminals of a VNO i in a beam j. Let MAXgi be the gateway-level maximum subscription of a VNO i, then the distribution of this VNO's maximum subscription on beam j can be calculated as follows: MAXi,j=(MAXgi*(alpha*Mi,j+(1−alpha)*Di,j) / (Summation of(alpha*Mi,j+(1−alpha)*Di,j) over all rays j=i to J) The present teachings enforce that the summation of the maximum bandwidth distribution in beams for a VNO does not exceed its maximum subscription at the gateway level. An exponential moving average (EMA) of Maxi,j can be calculated in each flow control epoch over multiple epochs within an update cycle (a few minutes), and at the time of the update, the last EMA value of Mi,j and Di,j can be used as the maximum subscription for each beam.

[0055] Once all maximum subscriptions of a VNO are distributed among eligible beams, the present teachings aggregate all of the VNO's maximum subscriptions, including any static maximum subscription for each beam. Based on the estimated current capacity, the oversubscription ratio is calculated in each beam. In the event of an oversubscription ratio imbalance across a beam, the maximum bandwidth distribution can be adjusted to keep the oversubscription ratio more or less balanced. The adjustment can also be made based on the VNO's priority. Embodiment 3: Distribution of a minimum subscription

[0056] A VNO may subscribe to a minimum bandwidth at the gateway or system level. In some embodiments, a gateway-level bandwidth subscription may be provided only on an outroute, only on an inroute, or both. For example, and for brevity, the present disclosure is directed to the outroute. The gateway-level bandwidth subscription may be used as guaranteed bandwidth for the VNO. The system may be dimensioned such that the bandwidth capacity of a system exceeds a total sum of the minimum bandwidths for all VNOs. In each flow control epoch, for example, every 100 milliseconds (ms), the GBM receives at least the subscription information of the aggregated minimum bandwidths of all active end devices separately for each beam and each VNO.The information can be collected by a gateway, received by a beam-level bandwidth manager, and provided by the beam-level bandwidth manager to the GBM. The GBM can execute a proportional fairness algorithm to partition a VNO's gateway-level minimum subscription into the gateway's subscribed beams based on the aggregate minimum subscriptions of the gateway's active end devices on each beam.

[0057] At startup or initialization, the GBM does not have information about the minimum bandwidth distribution across beams. The first distribution may be calculated during the first flow control cycle. In some embodiments, the distribution calculation may initially be performed in a faster space (less than a configured timer). After initialization, the GBM may calculate and update the minimum bandwidth distribution of each VNO across beams at each update cycle. The update cycle may be configurable, for example, with a default value of one (1) minute.

[0058] For minimum bandwidth distribution, the GBM can receive parameters from the CROs at each flow control cycle and, for example, generate an aggregated minimum subscription per VNO per beam across all active guaranteed or real-time sessions for end devices, such as adaptive constant bit rate (CBR) and committed information rate (CIR) end devices.

[0059] In embodiments, the GBM may update the minimum subscription distribution across beams every X minutes (on the order of minutes). However, the GBM may continuously calculate an exponential moving average of an aggregated minimum subscription at each flow control cycle as input parameters are received from CROs. In some embodiments, the GBM may be synchronized with the flow control cycle. If input messages are missing in any flow control cycle, the OBM may skip the calculation and resume the calculation in the next flow control cycle. After X minutes have elapsed, the GBM may use the last sample of the exponential moving average of the aggregated minimum subscription values ​​to calculate the minimum distribution. Example steps performed by the GBM may include: • Receiving demand messages from the beams • When the first flow control cycle is initialized (i.e., after the OBM starts), calculate the first distribution of the VNO's minimum subscription into subscribed beams. In some embodiments, store the aggregated minimum subscription values ​​per VNO per beam and start a bandwidth distribution timer. The bandwidth distribution timer can be set to 60 seconds or the like. • Perform the following until the bandwidth distribution timer expires for each flow control cycle (initially, an accelerated timer can be used as the bandwidth distribution timer, for example, the bandwidth distribution timer divided by 6). ◯ Calculate the new exponential moving average of the minimum subscription per VNO per beam, using the current aggregated minimum subscription per VNO per beam and the previous sample of the exponential moving average; ◯ Assigning a constant value to the ray if the exponential moving average on a ray is zero, where the constant value may be a percentage of a VNO's current aggregate minimum subscription across active terminals; and o Store the last exponential moving average of an aggregated minimum subscription per VNO per beam. • When the bandwidth distribution timer expires, calculate the minimum bandwidth distribution for subscribed beams for each VNO using an exponential moving average (EMA) of the aggregated minimum calculated for each VNO and for each beam. Distribute the minimum for a VNO across beams in a proportionally fair manner. For example, the minimum subscription on beam j for a VNO i can be calculated as (Gateway-level minimum subscription of VNO i * EMA of the aggregated minimum subscription of VNO i on beam j) / (Summation of the EMA of the aggregated minimum subscription of VNO i across all beams). • If a VNO cannot subscribe to a minimum subscription at the gateway level, the minimum subscription distribution will not be evaluated by the GBM • Once all minimum subscriptions of a VNO are distributed across eligible beams, aggregate the minimum subscription of each VNO, including any static minimum subscription for each beam. In embodiments, the aggregated minimum is always available across VNOs within a beam.

[0060] During a flow control cycle in which the inroute and outroute bandwidth managers receive a demand message or the bandwidth status report message containing aggregated minimum subscriptions from end devices, a total minimum subscription across all active adaptive CBR and CIR end devices of the i-th VNO on the j-th beam is calculated by the inroute and outroute bandwidth managers. Let K be the number of such end devices from a VNO i and on a beam j, and use the following variables: • MC kij = Configured minimum subscription of a terminal k for VNO i on beam j • TMCij=∑k=1KMCkij - Total minimum subscription of the i-th VNO on the j-th beam; K is the number of active ACBR and CIR terminals • MS i = Minimum subscription at global gateway level of VNO i • J = Number of beams supplied by a gateway

[0061] Then TMC can ijp as the total minimum subscription across active adaptive CBR and CIR terminals by the i-th VNO on the j-th beam in flow control cycle p, where the i-th VNO has subscribed to a minimum gateway-level bandwidth. In some embodiments, this calculation may be skipped for a VNO m if m does not subscribe to a minimum gateway-level bandwidth. If a VNO does not subscribe to a minimum gateway-level bandwidth, that VNO is not considered in calculating the minimum per-beam distribution.

[0062] An exponential moving average (EMA) for the aggregate minimum subscription in cycle p of VNO i on the j-th ray can be expressed as follows: • ETMC ijp = The exponential moving average of the total minimum subscription for the i-th VNO on the j-th ray at flow control cycle p • ETMC ijp-1 = The exponential moving average of the total minimum bandwidth subscription for the i-th VNO on the j-th beam at flow control cycle p-1 is: ETMCijp=α*TMCijp+(1−α)*ETMCijp−1 where α is the smoothing factor that determines the filter-equivalent time constant. If the allocation period is T, then the time constant τ of the EMA is equal to - T / ln(1 - α) (α = 0.1 corresponds to a time constant of 10 when T = 1 period). The default value of α is 0.1 and is configurable. The smoothing factor α and the time period T can be configured in the system. • If ETMCijp is zero, then it is assigned a constant value that is a percentage of a VNO's current aggregate minimum subscription across active endpoints. • MinD ij = Minimum subscription distribution of VNO i over beam j, where p is the last flow control cycle before the update: MinDij=MinSi∗ETMCijp∑j=1lETMCijp where MinSi = minimum subscription at global gateway level of VNO i

[0063] For each VNO, the aggregate minimum subscriptions of active Adaptive CBR and CIR QoS end devices are calculated for each beam to which a VNO subscribes. A proportional ratio is calculated for each VNO across beams using the total number of end devices per VNO per beam. For example, if a VNO has an aggregate minimum subscription of enabled ACBR / CIR end devices across three beams, and the aggregate number of these across three beams is 10 Mbps, 30 Mbps, and 60 Mbps, respectively, the ratio is 1:3:6 across three beams. If this VNO has a total minimum gateway-level subscription of 180 Mbps, the distribution would be 18 Mbps, 54 Mbps, and 108 Mbps, respectively.

[0064] The above procedure is performed for all VNOs in the system, and the minimum bandwidth per beam is obtained for all VNOs. In some embodiments, there may be some VNOs that subscribe to a static minimum bandwidth per beam in the system on the same beams where dynamic VNOs are present.

[0065] When the update interval timer X min expires, the last ETMC ijp used to calculate the minimum distribution values, known as MinD ij can be described.

[0066] The summation of the minimum bandwidth subscription across all VNOs within a beam is guaranteed to a VNO. This amount of bandwidth should generally always be available. Once all minimum subscriptions of a VNO are distributed across eligible beams, the process aggregates the minimum subscriptions of each VNO, including the static minimum subscriptions for each beam. Ideally, the aggregated minimum across VNOs within a beam is always available. Embodiment 4: Maximum subscription distribution

[0067] A VNO can subscribe to a maximum or peak bandwidth at the gateway level. This bandwidth can be treated as the best-effort bandwidth. In each flow control epoch, e.g., every 100 ms, the GBM can receive the aggregated maximum bandwidth subscription information of all active end devices separately for each beam and each VNO. In some embodiments, the GBM can also receive the current traffic demand per flow control epoch. The total maximum / peak subscription of each VNO in all beams can be based on the guaranteed maximum subscription of all active ACBR end devices and the maximum subscription of the CIR, BE, and TCBE end devices.

[0068] The GBM calculates a proportional fair average to distribute a VNO's gateway-level maximum subscription among its subscribed beams based on the aggregated maximum subscription of its active endpoints and the current demand from active endpoints on each beam. The steps of the process are captured as follows. In exemplary embodiments, the process is similar between an OBM and an IBM, except that the aggregated subscriptions can come from different locations and in different messages. Therefore, this section uses the term "Bandwidth Manager" to describe the process for both an OBM and an IBM, unless explicitly distinguished: • Receiving by the OBM demand messages from the beams and / or receiving by the IBM bandwidth status report messages from the gateways. • In some embodiments, when it is the first flow control cycle (i.e., after starting OBM and IBM), OBM and IBM calculate the first distribution of the maximum subscription from VNOs to subscribed beams in a proportionally fair manner based on the aggregated maximum subscription and bandwidth demand / usage. In some embodiments, bandwidth managers store the aggregated maximum subscription values ​​per VNO per beam and bandwidth demand / usage across end devices and start the bandwidth distribution timer, for example, 60 seconds.

[0069] For each flow control cycle, the bandwidth managers perform the following until the bandwidth distribution timer expires. Initially, an accelerated timer can be used, for example, the bandwidth distribution timer divided by 6. • Calculate the new exponential moving average of the maximum subscription and demand / usage per VNO per beam based on the current aggregated maximum subscription per VNO per beam, bandwidth demand / usage, and the previous sample of the exponential moving average; • Assigning a constant value to the new exponential moving average of the maximum subscription and demand / usage per VNO per beam if the exponential moving average is zero. In some embodiments, the constant value may be a percentage of the maximum subscription at the gateway level of a VNO; and • Store the last exponential moving average of the aggregated maximum subscription and demand / usage per VNO per beam.

[0070] When the bandwidth allocator timer expires, the bandwidth managers calculate the maximum bandwidth allocation for each VNO using the EMA (exponential moving average) of the aggregated maximum and bandwidth demand / usage calculated for each VNO and each beam. The bandwidth allocator distributes the maximum for a VNO proportionally and fairly across beams.

[0071] Once all maximum subscriptions of a VNO are distributed among admissible beams, the process aggregates all maximum subscriptions of each VNO, including any static maximum subscriptions for each beam. In embodiments, the aggregated maximum is oversubscribed across VNOs within a beam. Based on the estimated current capacity, the oversubscription ratio in each beam is calculated. In some embodiments, the results are stored and displayed, for example, to a network designer or operator.

[0072] During a flow control cycle in which the bandwidth managers receive a demand message or the bandwidth status report message containing aggregated maximum subscriptions from end devices, a total maximum subscription across active ACBR (maximum guaranteed), CIR (maximum / peak), TCBE, and BE (maximum / peak) end devices of the i-th VNO on the j-th beam is calculated by the in-route and out-route bandwidth managers. Let K be the number of such end devices from a VNO i and on beam j, and let • MxC kij = Configured maximum subscription of a terminal k (for ACBR, guaranteed maximum subscription, for CIR, TCBE and BE maximum or peak subscription) for VNO i on beam j • TMxCij=∑k=1KMxCkij - Total maximum subscription of the i-th VNO on the j-th beam; K is the number of active terminals • TD ij= Total demand / usage aggregated across all active terminals for a VNO i on beam j. On the inroute for ACBR terminals and within the guaranteed part of CIR, usage or throughput is used, and for TCBE, BE and the best effort part of CIR, demand or unstaffed backlogs are used • MxS i = Maximum subscription at global gateway level of VNO i • J = Number of beams served by a gateway • N = number of VNOs

[0073] An exponential moving average (EMA) for the aggregate maximum subscription at cycle p of VNO i on the j-th ray can be computed by the GBM, IBM, or OBM as: • ETMxC ijp = The exponential moving average of the total maximum subscription for the i-th VNO on the j-th ray at flow control cycle p • ETMxC ijp-1= The exponentially moving average of the total maximum bandwidth subscription for the i-th VNO on the j-th ray at flow control cycle p-1 . ETMxC ijp = α * TMxC ijp + (1 - α) * ETMxC ijp-1, where α is the smoothing factor that determines the filter-equivalent time constant. If the allocation period is T, then the time constant τ of the EMA is equal to - T / In(1 - α) (α = 0.1 is equivalent to the time constant 10 when T = 1 period). The default value of α can be set as a fraction of T, e.g., 0.1, and can be configurable. The smoothing factor α and the period T can be configurable in the system.

[0074] In some embodiments, an EMA for the aggregate demand and usage (aggregated across all traffic classes across all active terminals) may be calculated for a VNO and on each beam. The EMA for the aggregate demand at cycle p of VNO i on the j-th beam may be calculated as follows: • ETD ijp = The exponentially moving average of the total demand / usage for the i-th VNO on the j-th ray in the flow control cycle p • ETDC ijp-1 = The exponentially moving average of the total demand / usage for the i-th VNO on the j-th ray in flow control cycle p-1. • ETD ijp = α * TD ijp + (1 - α) * ETD ijp-1, where α is the smoothing factor that determines the filter-equivalent time constant. If the allocation period is T, then the time constant τ of the EMA is equal to - T / ln(1 - α) (α = 0.1 is equivalent to the time constant 10 when T = 1 period). The default value of α is 0.1 and is configurable. The smoothing factor α and the time period T can be configured in the system. • MaxD ij = Maximum subscription distribution of VNO i over beam j, where p is the last flow control cycle before the update MaxDij=MxSi∗(ETMxCijp∗beta+ETDijp∗(1−beta))∑j=1J(ETMxCijp∗beta+ETDijp∗(1−beta))

[0075] In the above calculation, beta can be a static factor that provides a deviation between a configured maximum and actual demand / usage. By default, the beta value can be adjusted to give more weight to actual demand / usage, for example, by setting beta to 0.3.

[0076] In embodiments, the process may also reduce the rollup of the maximum bandwidth distribution in beams for a VNO so that the rollup does not exceed the maximum subscription at the gateway level of a VNO.

[0077] The above process is performed for all VNOs in the system, and the maximum bandwidth per beam is obtained for all VNOs. In some embodiments, some VNOs may subscribe to a static maximum bandwidth per beam in the system on the same beams where dynamic VNOs are present. After calculating the static maximum bandwidth per beam, the system uses a capacity of each beam to calculate a predetermined maximum bandwidth value available to all VNOs on each beam.

[0078] Once all of a VNO's maximum subscriptions are distributed across eligible beams, the process adds up all of the VNO's maximum subscriptions, including any static maximum subscriptions for each beam. The process then calculates an estimated current capacity for each beam, which forms the basis for calculating the oversubscription ratio in each beam. In some embodiments, on the in-route side, the estimated capacity may be set as the effective capacity, which excludes in-route groups specifically assigned to one or more VNOs and in-route groups not used by any VNO. In the event of an oversubscription ratio imbalance across a beam, the maximum bandwidth distribution may be adjusted to keep the oversubscription ratio more or less balanced.

[0079] Once all maximum subscriptions of a VNO have been distributed among eligible beams, the algorithm adds up all of the VNO's maximum subscriptions, including any static maximum subscriptions for each beam. Based on the estimated current capacity, the oversubscription ratio is calculated in each beam. In the event of an oversubscription imbalance across a beam, the maximum bandwidth distribution is adjusted to keep the oversubscription ratio more or less balanced. The adjustment is also based on the VNO's priority.

[0080] Fig. 2 illustrates communication flows on an outroute chain according to various embodiments.

[0081] Per-gateway per VNO 202 demand messages are sent from a gateway to a beam-level bandwidth manager, e.g., OBM 204, 204', 204", in each flow control epoch, e.g., every 100 ms. The demand messages may be sent in response to a heartbeat message in the system. The OBM 204, 204', 204" may send demand messages to the GBM 206. These demand messages may include an aggregate demand request for each VNO per outroute beam if a beam serves multiple VNOs; otherwise, the demand request may include an aggregate demand request for only one VNO. The GBM 206 may calculate a per-beam bandwidth partition per VNO, which is provided to the OBM 204, 204', 204" to limit bandwidth consumption by the VNO on that beam.

[0082] Fig. 3 illustrates communication flows on an in-route chain according to various embodiments.

[0083] A beam-level bandwidth manager, for example, IBM 304, 304', 304", receives demand requests 302 from Very Small Aperture Terminals (VSATs) for active data flows from the VSATs. The IBM 304, 304', 304" periodically sends bandwidth status report messages per beam per VNO to a GBM 306, for example, every superframe period, such as every 360 ms interval. The GBM 306 can calculate a per-beam bandwidth partition per VNO to provide to the IBM 304, 304', 304" to limit bandwidth consumption by the VNO on that beam.

[0084] Fig. 4 illustrates a flowchart of an example method for bandwidth management for an ENT according to some embodiments.

[0085] A method 400 for bandwidth management for an ENT is disclosed. The method 400 may include an operation to configure the network 402. The method 400 may include an operation to provision multiple virtual network operators (VNOs) 404. The method 400 may include an operation to provision multiple communication beams 406. The method 400 may include an operation to assign subscriptions to VNOs 408.

[0086] The method 400 may include a beam-level bandwidth management operation 410. The beam-level bandwidth management operation 410 may be performed according to the procedures associated with the IBM and / or the OBM of Fig. 1. The method 400 may include an operation to allocate active terminals per beam per VNO 412. The method 400 may include an operation to aggregate the number of terminals by QoS per beam per VNO 413. The method 400 may include an operation to aggregate demand and subscriptions (maximum) of terminals per beam per VNO 414. The method 400 may include an operation to aggregate guaranteed demand and guaranteed subscription (minimum) per beam per VNO 416. The method 400 may include an operation to limit bandwidth consumption per VNO to the bandwidth distribution provided by the GBM 418.

[0087] The method 400 may include a global bandwidth management operation 420. The global bandwidth management operation 400 may be performed according to the methods associated with the GBM 140 of Fig.1. The method 400 may include an operation for receiving a number of terminals according to QoS per beam per VNO 421. The method 400 may include an operation for receiving demand and subscription (maximum) per beam per VNO 422. The method 400 may include an operation for receiving a guaranteed demand and a guaranteed subscription (minimum) per beam per VNO 424. The method 400 may include an operation for distributing the global subscription (maximum) of each VNO into subscribed beams per VNO 426. The method 400 may include an operation for distributing the guaranteed global (minimum) subscription of each VNO into subscribed beams per VNO 428. The method 400 may include an operation for proportionally balancing an oversubscription of the distribution (maximum) across a beam shared by the VNOs 430.The method 400 may include an operation to proportionally balance a.

[0088] Oversubscribing the guaranteed demand (minimum) across a beam shared by the VNOs 432. The method 400 may include an operation to provide an adjusted bandwidth distribution per beam per VNO to gateway 434.

[0089] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as exemplary forms for implementing the claims. Other configurations of the described embodiments are within the scope of this disclosure. Further, implementations consistent with the subject matter of this disclosure may include more or fewer acts than described, or implement acts in a different order than shown. Accordingly, the appended claims and their legal equivalents are intended to define only the invention, and not any of the specific examples provided.

Claims

[1] A method for bandwidth management for a host network operator (HNO), the method comprising: Providing shared beams shared by two or more of a plurality of Virtual Network Operators (VNOs), each VNO having a subscription comprising a global bandwidth limit applicable to a flow control epoch for each VNO and an end device subscription for each end device associated with the respective VNO; Aggregating a demand and the end-device subscriptions of active end-devices per shared beam per VNO in the current flow control epoch; Distributing the aggregate demand and subscriptions of active end devices for each shared beam per VNO into a distribution bandwidth per shared beam per VNO, where the distribution bandwidth is based on the respective global bandwidth limitation of the respective VNO; Proportionately balancing any oversubscription of the distribution bandwidths per shared beam per VNO based on proportionalities of the global bandwidth limitation of each VNO subscribing to the shared beam for each shared beam to provide a proportional distribution bandwidth per shared beam per VNO; and the subscription further includes a guaranteed minimum global bandwidth, a data flow is specified by the active end devices as either guaranteed or best-effort data flow, the aggregation further comprises aggregating the demand for and subscriptions of active terminals for data flows that are indicated as being guaranteed as one guaranteed demand and one guaranteed subscription of active terminals for each shared beam per VNO, and distributing further comprises setting the bandwidth distribution for each of the shared beams per VNO to be equal to or greater than the respective guaranteed demand and subscriptions of active terminals for each shared beam per VNO. [2] The method of claim 1, wherein distributing further comprises calculating an exponential moving average (EMA) of the aggregated demand and subscriptions of active terminals. [3] The method of claim 1, further comprising limiting bandwidth consumption by the active terminals of each shared beam in a flow control epoch occurring after the current flow control epoch based on the proportional distribution bandwidth per shared beam per VNO. [4] The method of claim 1, wherein the network comprises a satellite, a gateway, and a plurality of Very Small Aperture Terminals (VSATs), a plurality of communication beams comprise a radio frequency (RF) beam directed from the satellite to the gateway or to some of the VSATs, and the shared beams comprise some of the plurality of communication beams. [5] The method of claim 4, wherein at least one of the plurality of VSATs is mobile and capable of communicating using a first beam illuminating a first geographic area or a second different beam illuminating a second geographic area, the first beam and the second beam being selected from the shared beams, and the mobile VSAT moving from the first geographic area to the second geographic area. [6] The method of claim 1, further comprising: Changing a bandwidth capacity of one or more of the shared beams; and Limiting the distribution bandwidth or the proportional distribution bandwidth for each shared beam to be equal to or less than the respective bandwidth capacity of the shared beam. [7] The method of claim 1, wherein the shared beams are managed by a gateway operated by the ENT. [8] The method of claim 1, wherein the shared beams are managed by a plurality of gateways operated by the ENT. [9] The method of claim 1, further comprising proportionally compensating for oversubscription of the distribution bandwidths per shared beam based on the proportionalities of the global minimum bandwidth limit of each VNO subscribing to the shared beam for each shared beam to calculate the proportional distribution bandwidth per shared beam per VNO. [10] A system for performing bandwidth management for a host network operator (HNO), the system comprising: shared beams shared by two or more of a plurality of Virtual Network Operators (VNOs), each VNO having a subscription comprising a global bandwidth limit applicable to a flow control epoch for each VNO and an end device subscription for each end device associated with the respective VNO; a beam-level bandwidth manager for aggregating demand and end-device subscriptions of active end devices per shared beam per VNO in the current flow control epoch; and a Global Bandwidth Manager (GBM): to distribute the aggregate demand and subscriptions of active terminals for each shared beam per VNO into a distribution bandwidth per shared beam per VNO, where the distribution bandwidth is based on the respective global bandwidth limitation of the respective VNO; to proportionally compensate for any oversubscription of the distribution bandwidths per shared beam per VNO based on the proportionalities of the global bandwidth limitation of each VNO subscribing to the shared beam for each shared beam to provide a proportional distribution bandwidth per shared beam per VNO; and where the subscription includes a guaranteed minimum global bandwidth, a data flow is specified by the active parties as either guaranteed or best-effort data flow, the beam bandwidth manager is further operated to aggregate the demand for and subscriptions of active end devices for data flows that are indicated as guaranteed as one guaranteed demand and one guaranteed subscription of end devices for each shared beam per VNO, and the GBM is further operated to set the bandwidth distribution for each of the shared beams per VNO to be equal to or greater than the respective guaranteed demand and subscriptions of active terminals for each shared beam per VNO. [11] The system of claim 10, wherein the GBM for distributing is further operable to calculate an exponential moving average (EMA) of the aggregated demand and subscriptions of active terminals. [12] The system of claim 10, wherein the beam bandwidth manager is further operable to limit bandwidth consumption by the active terminals of each shared beam in a flow control epoch occurring after the current flow control epoch based on the proportional distribution bandwidth per shared beam per VNO. [13] The system of claim 10, wherein the network comprises a satellite, a gateway, and a plurality of Very Small Aperture Terminals (VSATs), a plurality of communication beams comprise a radio frequency (RF) beam directed from the satellite to the gateway or to some of the VSATs, and the shared beams comprise some of the plurality of communication beams. [14] The system of claim 13, wherein at least one of the plurality of VSATs is mobile and capable of communicating using a first beam illuminating a first geographic area or a second different beam illuminating a second geographic area, the first beam and the second beam being selected from the shared beams, and the mobile VSAT moving from the first geographic area to the second geographic area. [15] The system of claim 10, wherein the beam bandwidth manager is further operable to change a bandwidth capacity of one or more shared beams and to limit the distribution bandwidth or the proportional distribution bandwidth for each shared beam to be equal to or less than the respective bandwidth capacity of the shared beam. [16] The system of claim 10, wherein the shared beams are managed by a gateway operated by the ENT. [17] The system of claim 11, wherein the shared beams are managed by a plurality of gateways operated by the ENT. [18] The system of claim 11, wherein the GBM is further operable to proportionally compensate for oversubscription of the distribution bandwidths per shared beam based on the proportionalities of the global minimum bandwidth limit of each VNO subscribing to the shared beam for each shared beam to calculate the proportional distribution bandwidth per shared beam per VNO.

Images