VPN classification to reduce usage costs while retaining responsiveness

Abstract

Systems and methods for estimating the kind of traffic a VPN is carrying and determining which WAN connection to use for VPN network traffic to reduce usage costs while maintaining responsiveness in a VPN router or gateway. A VPN classifier examines VPN packets and estimates the type of traffic they are carrying and selects a policy for upstream or downstream traffic to determine which tunnel or tunnels to send network traffic. Further, policies may be applied to VPN packets as a function of usage knob.

Claims

1. A software defined wide area network (SD-WAN) device comprising: a first wide area network (WAN) connection with a first SD-WAN tunnel connected to the SD-WAN device; a second WAN connection with a second SD-WAN tunnel connected to the SD-WAN device; and a VPN classifier that examines encrypted packets of a virtual private network (VPN) connection, in which the packets are to be carried by the SD-WAN tunnels, estimates a traffic type of the packets of the VPN connection and sets a tunnel selection policy based at least partly on the traffic type for which tunnel of the first and second SD-WAN tunnels the packets of the VPN connection should be carried by; wherein: the VPN classifier uses recent measurements of packets of the VPN connection to estimate the traffic type based on a distribution of packets size ranges for packets transferred upstream and downstream; and the SD-WAN device selects from the first SD-WAN tunnel and second SD-WAN tunnel based on the tunnel selection policy.

2. The SD-WAN network device of claim 1, wherein the VPN classifier determines the distribution of packets in packet size ranges by counting sizes of upstream packets and downstream packets in a measurement period.

3. The SD-WAN device of claim 1, wherein the tunnel selection policy depends on an amount of higher-cost WAN connection carried historically.

4. The SD-WAN device of claim 1, wherein the recent measurements are on-going measurements taken periodically.

5. The SD-WAN device of claim 1, wherein the tunnel selection policy is determined using factors based on VPN connection upstream and downstream packet size and measurements of upstream and downstream packets.

6. The SD-WAN device of claim 5, wherein the factors include maximum packet inter-arrival time within a measurement period.

7. The SD-WAN device of claim 5, wherein the factors include a fraction of downstream traffic.

8. The SD-WAN device of claim 5, wherein the factors include a knob range of a usage knob, wherein the usage knob represents an urgency for reducing usage of the second WAN connection, and wherein the second WAN connection has a higher-cost per GB and lower-latency compared to the first WAN connection.

9. The SD-WAN device of claim 1, wherein the SD-WAN device is a SD-WAN router.

10. The SD-WAN device of claim 1, wherein the SD-WAN device is a SD-WAN gateway.

11. A software defined wide area network (SD-WAN) gateway that supports operation of one or more SD-WAN routers comprising: a first wide area network (WAN) connection with a first SD-WAN tunnel connected to the SD-WAN gateway; a second WAN connection with a second SD-WAN tunnel connected to the SD-WAN gateway; and a VPN classifier that examines encrypted packets of a virtual private network (VPN) connection, in which the packets are to be carried by the SD-WAN tunnels, estimates a traffic type of the packets of the VPN connection and sets a tunnel selection policy based at least partly on the traffic type for which tunnel of the first and second SD-WAN tunnels the packets of the VPN connection should be carried by; wherein: the VPN classifier uses recent measurements of packets of the VPN connection to estimate the traffic type, wherein the tunnel selection policy is updated on an on-going basis using rules with factors based on VPN connection upstream and downstream packet size and measurements of upstream and downstream packets, wherein the measurements by the VPN classifier include distribution of packets in packet size ranges determined by counting sizes of upstream packets and downstream packets in a measurement period; and the factors include a knob range of a usage knob, wherein the usage knob represents an urgency for reducing usage of the second WAN connection, wherein the second WAN connection has a higher-cost per GB and lower-latency compared to the first WAN connection.

12. The SD-WAN gateway of claim 11, wherein the tunnel selection policy is determined using factors based on VPN connection upstream and downstream packet size and measurements of upstream and downstream packets.

13. A method for assigning VPN network flow on a software defined wide area network (SD-WAN) device comprising: monitoring an encrypted VPN flow for a specified period of time, calculating traffic statistics of the VPN flow in the specified period of time; applying classification rules to the traffic statistics wherein the classification rules estimate a traffic type in the VPN flow; determining a tunnel selection policy based on matched classification rules with a highest score, wherein the tunnel selection policy depends on distribution of packets in packet size ranges for packets transferred upstream and downstream; and assigning VPN flow to a WAN interface based on the traffic type and the tunnel selection policy.

14. The method of claim 13 wherein the tunnel selection policy is determined using factors based on VPN connection upstream and downstream packet size and measurements of upstream and downstream packets, including determining distribution of packets in packet size ranges by counting sizes of upstream packets and downstream packets in a measurement period.

15. The method of claim 13, wherein the tunnel selection policy depends on an amount of higher-cost WAN connection carried historically.

16. The method of claim 13 wherein the calculated traffic statistics are taken when a specified number of bytes or packets have been accumulated.

17. The method of claim 13, wherein the classification rules include a maximum packet inter-arrival time within a measurement period.

18. The method of claim 13, wherein the classification rules include a fraction of downstream traffic.

19. The method of claim 13, wherein the classification rules include a setting of a usage knob, where the usage knob is a representative of an urgency for reducing usage of a WAN connection with a higher-cost per GB and lower-latency.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The drawing figures depict one or more implementations in accord with the present teachings, by way of example only, not by way of limitation. In the figures, like reference numerals refer to the same or similar elements. Furthermore, it should be understood that the drawings are not necessarily to scale.

(2) FIG. 1 illustrates a high-level block diagram of a system for VPN classification to reduce usage costs while retaining responsiveness.

(3) FIG. 2 illustrates a high-level block diagram of an SD-WAN router with a VPN classifier.

(4) FIG. 3 illustrates a high-level block diagram of an SD-WAN gateway with a VPN classifier.

(5) FIG. 4 illustrates upstream packet flow through the SD-WAN router.

(6) FIG. 5 illustrates upstream packet flow through the SD-WAN gateway.

(7) FIG. 6 illustrates downstream packet flow through the SD-WAN gateway.

(8) FIG. 7 illustrates downstream packet flow through the SD-WAN router.

(9) FIG. 8 is a sequence diagram illustrating a method for VPN classification to reduce usage costs while retaining responsiveness in accordance with this disclosure.

(10) FIG. 9 illustrates a block diagram showing an example of router hardware upon which aspects of this disclosure may be implemented.

(11) FIG. 10 illustrates a block diagram showing an example computer system upon which aspects of this disclosure may be implemented.

DETAILED DESCRIPTION

(12) In the following detailed description, numerous specific details are set forth by way of examples to provide a thorough understanding of the disclosed subject matter. It may become apparent to persons of ordinary skill in the art, though, upon reading this disclosure, that one or more disclosed aspects may be practiced without such details. In addition, description of various example implementations according to this disclosure may include referencing of or to one or more known techniques or operations, and such referencing can be at relatively high-level, to avoid obscuring of various concepts, aspects and features thereof with details not particular to and not necessary for fully understanding the present disclosure.

(13) This disclosure describes systems and methods for estimating the kind of traffic a VPN is carrying and determining which WAN connection to use for VPN network traffic to reduce usage costs while maintaining responsiveness. A VPN classifier examines VPN packets and estimates the type of traffic they are carrying and selects an SD-WAN policy (also referred to herein as a “policy” or “selection policy”) for the upstream or downstream VPN packets. The policy determines, together with current WAN status, which tunnel or tunnels thru which to send the VPN packets. Further, policies may be applied to VPN packets as a function of a usage knob as described herein.

(14) FIG. 1 illustrates a high-level block diagram of a system 100 for VPN classification to reduce usage costs while retaining responsiveness. The system 100 operates in a satellite/wireless hybrid internet access setup where end user traffic, as dictated by the appropriate policies, may be carried between the SD-WAN Router 110 and SD-WAN gateway 112 via a combination of a satellite access network and a wireless access network as described below. The system 100 uses VPN classifiers 114, 116 with VPN classification rules 118 to dynamically adjust the policy applied to a VPN connection to more appropriately trade off the lower-cost per GB and higher thruput of a satellite metered internet connection with the lower latency but higher cost per GB of a wireless metered internet connection to provide a more responsive end-user experience than possible with a single satellite Internet connection at a lower-cost with higher bulk-transfer thruput than possible with a single wireless Internet connection.

(15) Again referring to FIG. 1, the system 100 includes an SD-WAN router 110 which provides internet access for a remote host 120 over a remote local area network (LAN) 122 over a connection 124. In the illustrated implementation, the SD-WAN router 110 communicates with the SD-WAN gateway 112 over two WAN connections. The first WAN connection includes a satellite modem 126 that supports a satellite tunnel 128 through a satellite access network 130 to the SD-WAN gateway 112. The second WAN connection includes a wireless modem 132 that supports a wireless tunnel 134 through a wireless LTE access network 136 to the SD-WAN gateway 112. The WAN connections support network communication between the remote host 120 and a public internet host 140 over the internet 138. In the illustrated implementation the two WAN connections are a satellite connection and an LTE access network. Other alternatives are expressly contemplated. For example, in an alternative implementation a LEO (Low-Earth Orbit) satellite internet connection or a 5G wireless transport may be utilized in place of the LTE. Any combination of WAN connections may be implemented that support network communication controlled by the SD-WAN router and SD-WAN gateway as described herein.

(16) FIG. 2 illustrates a high-level block diagram of the SD-WAN router 110. The interconnection and operation of the blocks within the SD-WAN router 110 are described further below with reference to FIGS. 4 and 7. The SD-WAN router 110 utilizes a VPN classifier 114 to help set the policy for an Upstream VPN IP flow. The SD-WAN Router 110 includes a LAN interface 210 that allows it to exchange packets with multiple LAN host devices (with only one shown in the figure, in this case a VPN Client 212). The classifier 214 examines packet IP addresses, ports and other packet characteristics to identify IP flows (upstream and downstream). The VPN classifier 114 examines VPN packets and estimates the type of traffic they are carrying and selects SD-WAN policies for each upstream VPN packet. The SD-WAN policy block 216 determines which tunnel or tunnels through which to send an upstream packet based on an upstream packet's tagged policy. The SD-WAN resequencer 218 resequences an IP flow's received downstream packets. The satellite tunnel end-point 220 encapsulates upstream packets in another packet, a satellite tunnel packet, and de-encapsulates received satellite tunnel packets. The satellite WAN interface 222 sends and receives packets via a satellite modem 126 on the satellite tunnel 128 over the satellite access network 130 (FIG. 1) and onto the public internet 138. The wireless tunnel end-point 226 encapsulates upstream packets in another packet, a satellite tunnel packet, and de-encapsulates received satellite tunnel packets. The wireless WAN interface 228 sends and receives packets via the wireless modem 132 on the wireless tunnel 134 over the wireless access network 136 (FIG. 1) and on to the public internet 138. The SD-WAN router 110 may further include a usage knob 230 that is used by the VPN classifier 114 as described further below.

(17) FIG. 3 illustrates a block diagram of an implementation of the SD-WAN gateway 112. The interconnection and operation of the blocks within the SD-WAN gateway 112 are described further below with reference to FIGS. 5 and 6. The SD-WAN gateway 112 utilizes a VPN classifier 116 to set the policy for a downstream VPN IP flow. The organization of an SD-WAN gateway 112 is similar, but not identical, to that of the SD-WAN Router 110 described above. A significant difference is that an SD-WAN gateway 112 typically supports communications with multiple SD-WAN routers while an SD-WAN router often communicates with a single SD-WAN Gateway (or a small number of SD-WAN Gateways). The SD-WAN gateway 112 includes a network interface 310 through which it sends and receives satellite and wireless tunnel packets via the public internet 138. In the implementation shown, this same network interface 310 also sends and receives packets to public internet servers (in this case a VPN gateway 312) via the public internet. In another implementation, an SD-WAN Gateway has a separate Network Interface that sends and receives un-tunneled packets to their upstream host(s). This is useful when the SD-WAN is supporting a private network. The classifier 314 examines packet IP addresses, ports and other packet characteristics to identify IP flows (upstream and downstream). The VPN classifier 116 examines VPN packets and estimates the type of traffic they are carrying and selects the SD-WAN policy for each downstream VPN packet. The policy object 316 determines which tunnel or tunnels through which to send a downstream packet based on a downstream packet's tagged policy. The SD-WAN resequencer 318 resequences an IP flow's received upstream packets. The satellite tunnel end-point 320 encapsulates downstream packets in another packet, a satellite tunnel packet, and de-encapsulates received satellite tunnel packets. The wireless tunnel end-point 322 encapsulates downstream packets in another packet, a satellite tunnel packet, and de-encapsulates received satellite tunnel packets. The SD-WAN gateway 112 may further include a usage knob 324 that is used by the VPN classifier 116 as described further below.

(18) FIGS. 4 and 5 illustrate upstream packet flow through the system 100 described above. FIG. 4 illustrates the upstream packet flow through the SD-WAN router 110. In the illustrated implementation, the SD-WAN router 110 processes upstream VPN packets as shown by the arrows from the VPN client 212. The VPN client 212 sends packets to its VPN gateway via the SD-WAN Router 110. The LAN interface 210 (typically with its TCP/IP stack) passes the packets into the classifier 214 which identifies packets as being part of a VPN connection. The classifier 214 passes the VPN packets thru the VPN classifier 114 which measures the packet and, based on the measurement of the VPN connection's upstream and downstream packets estimates the type of traffic currently being carried by the VPN connection. In the illustrated implementation, the Classifier 214 tags upstream packets which are not VPN packets itself with the appropriate policy and forwards those packets to the SD-WAN policy block 216 without passing them thru the VPN classifier 114. The VPN classifier 114, based on the estimated traffic type and the usage knob setting, tags the packet with the SD-WAN policy to be applied to the packet and forwards the packet back to the classifier 214. In an alternative implementation, the VPN classifier 114 forwards the tagged packet to the SD-WAN policy block 216 directly without passing it back thru the classifier 214. The classifier 214 forwards the packet to the SD-WAN policy block 216 which, based tagged policy and on the current measured characteristics of the satellite WAN connection and the wireless WAN connections forwards the packet into the satellite tunnel end-point 220 or the wireless tunnel end-point 226 or both. The satellite tunnel end-point 220 encapsulates the packet, when the policy calls for it to be sent via satellite, and forwards it thru the satellite WAN interface 228 thru to the satellite modem 126 and its Metered ISP on the satellite tunnel 128 thru the public internet 138 to the SD-WAN gateway 112 (FIG. 1). Examples of such encapsulation techniques include Generic Routing Encapsulation (GRE) and IPSec tunneling. The wireless tunnel end-point 226 encapsulates the packet, when the policy calls for it to be sent via wireless, and forwards it thru the wireless WAN interface 228 thru the wireless modem 132 and its Metered ISP on the wireless tunnel 134 thru the public internet 138 to the SD-WAN Gateway 112 (FIG. 1).

(19) FIG. 5 illustrates the upstream packet flow through the SD-WAN gateway 112. The SD-WAN Gateway's processing of an upstream tunneled VPN packet, in the preferred embodiment is as follows. Packets are received over the wireless tunnel 134 and the satellite tunnel 128 on the public internet as described above and delivered to the network interface 310 based on the outer packet header's destination address. The network interface 310 (and typically its TCP/IP stack) delivers the packet, when it is being carried by the satellite tunnel 128, to the satellite tunnel end-point 320 which de-encapsulates the packet and forwards it to the SD-WAN resequencer 318. The network interface 310 (and typically its TCP/IP stack) delivers the packet, when it is being carried by the wireless tunnel 134, to the wireless tunnel end-point 322 which de-encapsulates the packet and forwards it to the SD-WAN resequencer 318. The SD-WAN resequencer 318, on a per IP-flow basis, puts an IP flow's packets back in sequence when practical and forwards packets immediately when they are in-sequence or with sequence gaps (or potentially out of sequence) when they have been buffered waiting to become in sequence for a seemingly long enough period of time. The SD-WAN resequencer 318 forwards the packet to the classifier 314 which identifies the packet as being part of a VPN connection. The classifier 314 passes VPN packets, of which this is one, to the VPN classifier 116. In this implementation, the classifier 314 bypasses the VPN classifier 116 and forwards upstream packets which are not VPN packets directly out the network interface 310 on their way to their addressed destination. The VPN classifier 116 measures the packet, as part of measuring the VPN connection's behavior, and forwards it back thru the classifier 314. In an alternative implementation the VPN classifier 116 forwards the upstream VPN packet to the network Interface 310 directly without passing it back thru the classifier 314. The classifier 314 forwards the packet thru the network interface 310 (and typically its IPSec Stack). The network interface 310 forwards the packet on its way thru the public internet on its way to the VPN gateway 322.

(20) FIGS. 6 and 7 illustrate downstream packet flow through the system 100 described above. FIG. 6 illustrates the downstream packet flow through the SD-WAN gateway 112. The SD-WAN Gateway's 112 processing of a downstream VPN packet from the VPN gateway 322 to the VPN client 212 (FIG. 7) in the preferred embodiment, is as follows. First, the VPN gateway 322 sends packets over the public internet 138 to the network interface 310 of the SD-WAN gateway 112. This SD-WAN gateway processes packets to send to SD-WAN router 110 either by performing a Network Address Port Translation (NAPT) operation on the untunneld VPN packets thus causing the VPN gateway 322 to send downstream packets via the SD-WAN gateway 112 or by the SD-WAN gateway 112 advertising a route to the VPN Client. The network interface 310 (typically with its TCP/IP stack) passes to the classifier 314 downstream packets which are not tunneled packets from a SD-WAN router 110 and which are destined to be carried by a tunnel downstream to a SD-WAN router 110. The classifier 314 identifies a downstream VPN packet as being part of a VPN connection and passes the packet thru the VPN classifier 116. In the illustrated implementation, the classifier 314 tags downstream packets which are not VPN packets itself with the appropriate policy and forwards those packets to the SD-WAN policy block 312 without passing them thru the VPN classifier 116. The VPN classifier 116 measures the packet and, based on the measurement of the VPN connection's upstream and downstream packets estimates the type of traffic currently being carried by the VPN connection. The VPN classifier 116, based on the estimated traffic type and the usage knob setting, tags the packet with the SD-WAN policy to be applied to the packet and forwards the packet back to the classifier 314. In an alternative implementation, the VPN classifier 116 forwards the tagged packet to the SD-WAN policy block 312 directly without passing it back thru the classifier 314. The classifier 314 forwards the packet to the SD-WAN policy block 312 which, based on the tagged policy and current measured characteristics of the satellite WAN connection and the wireless WAN connections forwards the packet into the satellite tunnel end-point 320 or the wireless tunnel end-point 322 or both. The satellite tunnel end-point 320 encapsulates the packet, when the policy calls for it to be sent via satellite, and forwards the encapsulated packet thru the network on its way thru the public internet and back thru the satellite metered ISP to the SD-WAN router 110 (FIG. 7). Examples of such encapsulation techniques include Generic Routing Encapsulation (GRE) and IPSec tunneling. The wireless tunnel end-point 322 encapsulates the packet, when the policy calls for it to be sent via wireless, and forwards the encapsulated packet thru the network on its way thru the public internet and back thru the wireless metered ISP to the SD-WAN router 110.

(21) FIG. 7 illustrates the downstream packet flow through the SD-WAN router 110. The SD-WAN Router's processing of a downstream tunneled VPN packet, in the illustrated implementation is as follows. The public internet 138 carries such a packet and routes it thru either the satellite or wireless metered ISP and delivers it to the satellite WAN interface 222 or wireless WAN interface 228 based on the outer packet header's destination address. The satellite WAN interface 222 (and typically its TCP/IP stack) delivers the packet, when it is being carried by the satellite tunnel 128, to the satellite tunnel end-point 220 which de-encapsulates the packet and forwards it to the resequencer 218. The wireless WAN interface 228 (and typically its TCP/IP stack) delivers the packet, when it is being carried by the wireless tunnel 134, to the wireless tunnel end-point 226 which de-encapsulates the packet and forwards it to the SD-WAN resequencer 218. The resequencer 218, on a per IP-flow basis, puts an IP flow's packets back in sequence when practical and forwards packets immediately when they are in-sequence or with sequence gaps (or potentially out of sequence) when they have been buffered waiting to become in sequence for a seemingly long enough period of time. The resequencer 218 forwards the packet to the classifier 214 which identifies the packet as being part of a VPN connection. The classifier passes VPN packets, of which this is one, to the VPN classifier 114. The Classifier bypasses the VPN classifier 114 and forwards downstream packets which are not VPN packets directly out the LAN interface on their way to their addressed LAN host. The VPN classifier 114 measures the packet, as part of measuring the VPN connection's behavior, and, in the illustrated implementation, forwards it back thru the classifier 214. Alternatively, in another implementation the VPN classifier 114 may forward the downstream VPN packet to the LAN interface 210 directly without passing it back thru the classifier 214. The classifier 214 forwards the packet thru the LAN interface 210 (and typically its IPSec Stack). The LAN interface 210 forwards the packet on its way across the LAN on its way to the VPN client 212.

(22) FIG. 8 is a flow diagram of a process or method 800 for VPN classification to reduce usage costs while retaining responsiveness in accordance with this disclosure. Operations in the example method 800 could be performed by various entities in system 100. This flow diagram is only for purposes of example and is not intended as a limitation regarding implementations and practices in accordance with this disclosure. First, the system monitors VPN flow for a specified period of time, for example about 500 ms (step 810). The system then calculates traffic statistics of the VPN flow in the specified time period (step 820). Next, the system applies classification rules to the traffic statistics (step 830). The system then determines a SD-WAN selection policy based on matched classification rules with the highest score (step 840) and then assigns VPN flow to a WAN interface based on the VPN policy (step 850).

(23) VPN Classifier Packet Measurements And Calculations

(24) As has been discussed, the VPN classifiers 114, 116 examine and measure VPN connections upstream and downstream packets. The term VPN classifier as used herein refers to both the upstream VPN classifier 114 and the downstream VPN classifier 116. As will be discussed in an upcoming section, these measurements are restarted on an on-going basis (typically on a periodic basis) where the time between restarts is referred to as a measurement period. For each upstream packet, the VPN classifier 114 performs the following steps.

(25) Add the packet size to an upstream packet size total since the start of the current measurement period thereby allowing a bit rate to later be calculated.

(26) Increment the count of upstream packets since the start of the current measurement period thereby allowing a mean-packet size to later be calculated.

(27) Maintain a set of packet size ranges with the count of upstream packets in that range since the start of the current measurement period and increment the count of packets for packet size range that the packet's size falls into. The count of the number of packets in various ranges is a mechanism for measuring the distribution of packet sizes or packet size distribution. In one implementation, there are 8 buckets with the ranges from 0 . . . 199, 200 . . . 399 and so on to packet size of 1400 bytes or greater. This allows, for example, the fraction of large and small packets to be later be calculated.

(28) Measure the packet-interarrival time (the time since a previous of this IP flow's packet was received) and count the number of such packet inter-arrival times that are less than a configurable threshold since the start of the current measurement period.

(29) Measure the packet-interarrival time (the time since a previous of this IP flow's packet was received) and keep the maximum such interarrival time since the start of the current measurement period. For each downstream packet, the VPN classifier 116 does the same only for inbound packet measurements.

(30) VPN Classifier Periodic Calculations

(31) The VPN Classifier, in the illustrated implementation, on an on-going basis and preferably on a periodic basis, totals up the per-packet measurements and calculations to help characterize the traffic the VPN connection is carrying. The period of time between such calculations is referred to as a measurement period and the totaling up of measurements is based on the packets received during that measurement period. In the preferred embodiment the measurement period is 5 seconds long. The VPN Classifier computes (upstream and downstream) the following for each such a measurement period.

(32) Thruput (aka Bit rate)—computed from the duration of the period and the total size of the upstream or downstream packets.

(33) Packets-Per-Second (PPS)—computed from the duration of the period and the number of packets processed.

(34) Number of packets in each packet size range.

(35) Maximum period of time between packets (measured in milliseconds in the preferred embodiment).

(36) Number of packets which arrive with an interpacket arrival time (the time since the previous packet) less than a configurable threshold, default of 100 milliseconds.

(37) VPN Classifier Periodic Traffic Policy General Principles

(38) The VPN Classifier, in the illustrated implementation, updates an outbound VPN IP flow's policy once every measurement period. The update is based on the current and recent measurement period measurements and the usage knob setting. The usage knob setting is generally based on how much LTE usage is to be reduced where a higher knob setting indicates a higher desired level of LTE usage reduction. The usage knob in the preferred embodiment is set based on the LTE usage month-to-date as it is compared to a threshold monthly limit and how much of the month remains. The general principles guiding the policy selection are as follows: As the usage knob increases, the policy progressively shifts more traffic to satellite starting with downstream traffic. Downstream traffic is shifted first because it has lower latency than upstream traffic which, due to upstream bandwidth reservation techniques, may involve an extra satellite round-trip. When minimizing LTE use is important, all of the VPN's traffic in a given direction should go over satellite when any of its traffic should go over satellite. This is because satellite capacity is sufficient to handle a VPN connection's traffic and therefore if the traffic is split between satellite and LTE during resequencing of the satellite and LTE packets all of those packets are impacted by satellite's higher latency. Another general principle is that VPN traffic is latency sensitive and should be given a high priority when the underlying SD-WAN tunneling mechanism includes support for traffic prioritization or providing Quality Of Service even when a VPN IP flow is set to be carried exclusively by satellite. In the preferred embodiment, for flexibility and adjustability, the categorization of traffic is based on a set of configurable rules where, at any given time, more than one rule may “match” recent traffic and the current usage knob setting and where each such rule has a configured score and where the highest-scoring matching rule categorizes the traffic and thus selects the IP flow's policy.

(39) Detected Traffic Types and their Handling

(40) The VPN Classifier's classification rules provide a means for estimating traffic of the types that are identified in the table that follows. Table 1 shows some examples of VPN traffic types. The Usage Urgency column qualitatively indicates what the usage knob setting is calling for in terms of how aggressively usage should be shifted to Satellite and is one or more of the following: Low, Med (medium) and High. The resulting policy is shown in the table's Preferred Transport are as follows: SATELLITE—utilize the satellite Internet connection when it is working well and fall-back to wireless when not and when wireless is working well. WIRELESS—utilize the wireless Internet connection when it is working well and fall-back to satellite when not and when satellite is working well.

(41) TABLE-US-00001 TABLE 1 • Traffic Usage Preferred Type Direction Urgency Transport Description Deep Idle Up & Dn Low SATELLITE When traffic has been consistently low a Up & Dn Med & High SATELLITE longer time. Minimize WIRELESS usage. Light Idle Up Low WIRELESS When traffic has been consistently low for a short period of time. Minimize upstream latency. Dn Low SATELLITE Save WIRELESS usage Up & Dn Med & High SATELLITE Minimize WIRELESS usage Voice Only Up & Dn Low WIRELESS A voice only person-to-person or Call conference call. Up Med WIRELESS Dn Med SATELLITE Save bandwidth on the lower-latency Satellite downstream. Up & Dn High SATELLITE Save bandwidth even though the delay will be very noticeably bad. Audio Conf Up & Dn Low WIRELESS A conference call where the user's Call Mic Microphone is muted. Low Urgency == Muted best service. Up & Dn Med & High SATELLITE Minimize WIRELESS usage since responsiveness is not needed because the user is not talking. Audio Conf Up & Dn Low WIRELESS A conference call where the user's Call Mic Microphone is not muted. Low Urgency == Unmuted best service. Up Med WIRELESS Maintain responsiveness. Dn Med SATELLITE Give up some responsiveness Up & Dn High SATELLITE Give up responsiveness to reduce LTE usage. Video Conf Up & Dn Low WIRELESS A conference call where the user's Call Mic Microphone is muted. Low Urgency == Muted best service. Up & Dn Med & High SATELLITE Minimize WIRELESS usage since responsiveness is not needed. Video Conf Up & Dn Low WIRELESS A conference call where the user's Call Mic Microphone is not muted. Low Urgency == Unmuted best service. Up Med WIRELESS Maintain responsiveness. Dn Med SATELLITE Give up some responsiveness Up & Dn High SATELLITE Give up responsiveness to reduce LTE usage. Bulk Dnload Up Low & Med WIRELESS Minimize ack delay to increase thruput Dn Low & Med SATELLITE Minimize WIRELESS usage. Up& Dn High SATELLITE Minimize WIRELESS usage. Bulk Upload Up Low & Med SATELLITE Minimize WIRELESS usage. Dn Low & Med WIRELESS Minimize ack delay to increase thruput Up& Dn High SATELLITE Minimize WIRELESS usage. Other Up & Dn Low WIRELESS Maximize responsiveness Up Med WIRELESS Keep some WIRELESS responsiveness Dn Med SATELLITE Reduce WIRELESS usage Up & Dn High SATELLITE Minimize WIRELESS usage

(42) VPN Classification Rule Format

(43) The VPN classifier receives a set of VPN Classification rules that are used to estimate traffic type and, as a function of usage knob setting, select the SD-WAN policy to be applied to an outbound IP flow. Each rule, in the preferred embodiment, has the following mandatory and optional parameters that define the actions to be taken when the rule matches recent measurement period measurements:

(44) name (<alphanumeric+‘_’>)—REQUIRED, for documentation, tracing and logging purposes.

(45) description(<string>’)—REQUIRED, explains for documentation purposes what kind of traffic the rule is intended to match and what the resulting policy should be.

(46) score (<dec>)—REQUIRED, this rule's score where the highest matching score's action criteria are put in place.

(47) knobrange (<dec>,<dec>)—OPTIONAL with a default of 0, 100. This rule only applies when the LTE usage knob is within the range.

(48) upcontractid (<alphanumeric+‘_’>˜ . . . )—OPTIONAL, overrides the default upstream performance contract for this flow when the rule is the high-scoring match. May have multiple ‘˜’ separated contract IDs where the first applies to knob level one, the second to knob level two and so on where the last applies to its knob level and all higher knob levels. NOTE: The upcontractid field is used by the SD-WAN Router and the SD-WAN Router ignores rules that do not have an upcontractid parameter. NOTE: A performance contract defines an SD-WAN policy which may be as simple as always use one WAN or the other or may define various characterizations of expected performance (thruput capacity, latency, jitter, etc.) to guide the WAN selection. A typical rule's performance contract specifies a preferred WAN provided it is operating reasonably well compared to the other WAN.

(49) dncontractid (<alphanumber+‘_’>˜ . . . )—OPTIONAL, overrides the default dnstream performance contract for this flow when the rule is the high-scoring match. May have multiple ‘˜’ separated contract IDs where the first applies to knob level one, the second to knob level two and so on where the last applies to its knob level and all higher knob levels. NOTE: The dncontractid field is used by the SD-WAN Gateway and the SD-WAN Gateway ignores rules that do not have an upcontractid parameter.

(50) Each rule has a set of match criteria which are used to determine, at the end of a measurement period, whether the rule has matched recent measurement period measurements.

(51) multiperiodmatch (<dec N>˜<dec M>)—OPTIONAL, with a default of 1, 1 (matches on any single matching reading). This rule only matches when at least, N of the last M measurement periods otherwise matched this rule.

(52) or(<flag>˜<flag> . . . )—OPTIONAL, when present, instead of requiring that all of the specified ranges match, all of the non-default ranges must match except those listed in this parameter's values where at least one of those listed in this parameter must match

(53) percentdnstreamrange(<dec>,<dec>)—OPTIONAL with a default of 0, 100 (matching any percentage of downstream traffic). This refers to the percent of traffic (in bytes) downstream during the measurement period.

(54) upmaxinterpacketmsrange (<dec>˜<dec>)—OPTIONAL, with a default of 0 and very large value thereby matching virtually any value.

(55) upsmallinterpacketmspercentrange—OPTIONAL, with a default of 0 and 100 thereby matching with a default of virtually any value. Gives the percentage of upstream packets which arrive less than a configurable threshold ms (default=100) from the previous packet.

(56) upthruputrange (<dec>˜<dec>)—OPTIONAL, with a default of 0 and a very large value thereby matching any upstream thruput value. Gives the range of acceptable upstream thruput for the measurement period in units of kilobits/sec.

(57) upppsrange(<dec>˜<dec>)—OPTIONAL has the upstream packets per second range with a default of 0 and a very large value thereby matching upstream packet per second measurement.

(58) uppacketsizerangelist (<list entry>;<list entry> . . . ) OPTIONAL with one or more ‘;’ delimited list entries. The list entries are formatted as: (<hex>˜<dec>˜<dec>) where the hex is a hexadecimal bit map with 8 bits with each bit corresponding to one of the packet size buckets, first decimal is the low end of the range and the second decimal is the high end of the range.

(59) uppercentsmallpktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets from 0 to 399 bytes in length.

(60) uppercentlargepktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets >1200 bytes in length.

(61) uppercentverylargepktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets >1400 bytes in length.

(62) dnmaxinterpacketmsrange (<dec>˜<dec>)—OPTIONAL, with a default of 0 and very large value thereby matching virtually any value.

(63) dnsmallinterpacketmspercentrange—(<dec>˜<dec>) OPTIONAL, with a default of 0 and 100 thereby matching virtually any value. Gives the percentage of downstream packets which arrive less than a configurable threshold ms (default=100) from the previous packet.

(64) dnthruputrange (<dec>˜<dec>)—OPTIONAL, with a default of 0 and a very large value thereby matching any dnstream thruput value. Gives the range of acceptable downstream thruput for the measurement period in units of kilobits/sec.

(65) dnppsrange(<dec>˜<dec>)—OPTIONAL has the downstream packets per second range with a default of 0 and a very large value thereby matching downstream packet per second measurement.

(66) dnpacketsizerangelist (<list entry>;<list entry> . . . ) OPTIONAL with one or more ‘;’ delimited list entries. The list entries are formatted as: (<hex>˜<dec>˜<dec>) where the hex is a hexadecimal bit map with 8 bits with each bit corresponding to one of the packet size buckets, first decimal is the low end of the range and the second decimal is the high end of the range.

(67) dnpercentsmallpktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets from 0 to 399 bytes in length.

(68) dnpercentlargepktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets >1200 bytes in length.

(69) dnpercentverylargepktrange(<dec>,<dec>)—OPTIONAL with a default of 0 to 100, a convenient way of configuring an uppacketsizerangelist entry for packets >1400 bytes in length.

(70) VPN Classification Rule Examples

(71) The following tables provide example classification rules for common traffic types. These rules may evolve over time as traffic types change and as a better understanding of the characteristics of various traffic types is accumulated. The set of rules to be used are typically arrived at empirically, that is, by generating an example type of traffic and examining the traffic characteristics and selecting a combination of match criteria peculiar to that traffic. The set of rules can be customized as needed to detect specific types of traffic including, for example, the traffic characteristics of different vendors' conferencing applications.

(72) The first rule shown in Table 2 defines a WAN selection policy to be used the traffic is deemed to be “idle” and that no end-user is expected to be actually waiting for something to complete.

(73) TABLE-US-00002 TABLE 2 Parameter Value Explanation Name Idle Provides the name for this rule. description “Identify situations where Describes the type of traffic the rule traffic indicates that no human is carrying and the desired policy for has been waiting for operations that kind of traffic. to complete. Idle traffic should be carried with minimum cost regardless of usage knob setting.” Score 1100 A relatively low score so that other rule matches will take precedence should they occur. knobrange 1~6 Use this rule regardless of knob setting. upcontractid preferSatellite The policy is to use the satellite to carry this traffic provided it is working well. dncontractid preferSatellite The policy is to use the satellite to carry this traffic provided it is working well. multiperiodmatch 1~2 The rule matches when either the current measurement period or the previous measurement period matched. Or dnmaxinterpacketmsrange~ All of these match criteria are upmaxinterpacketmsrange~ deemed to have matched when any dnppsrange~ upppsrange of these match criteria match. dnppsrange 0~1 Match on downstream packets per second between 0 and 1 inclusive. upppsrange 0~1 Match on upstream packets per second between 0 and 1 inclusive. dnmaxinterpacket     800~2000000000 Match when the maximum msrange downstream packet inter-arrival time is at least 800 ms. upmaxinterpacket     800~2000000000 Match when the maximum upstream msrange packet inter-arrival time is at least 800 ms.

(74) A second rule shown in Table 3 defines the policy to be used the traffic is that of a bulk upload (for example sending and email with a large picture attached).

(75) TABLE-US-00003 TABLE 3 Parameter Value Explanation Name BulkUpload Provides the name for this rule. description “Traffic that is dominated by Describes the type of traffic the rule high-speed and large packets is is carrying and the desired policy for deemed to be an upload which that kind of traffic. should be carried via satellite. When the usage knob is not elevated the acknowledgement traffic (downstream traffic) is carried wireless which should allow for a higher-speed upload.” Score 100000 A relatively high score so that other rule matches will not take precedence should they occur. knobrange 1~6 Use this rule regardless of knob setting. upcontractid prefers at The policy is to use the satellite to carry this traffic provided it is working well. dncontractid preferWireless, preferSatellite The policy is to use the wireless when wireless usage has been consistently low (usage knob level 1) and to use satellite otherwise to conserve the more expensive wireless usage. multiperiodmatch 1~1 The rule matches when the current measurement period matched. upthruputrange     300~2000000000 Match on high upstream thruput (above 300 kbps). dnthruputrange  10~500 Match on low to moderate downstream thruput. Uppercentvery  90~100 Match when upstream packets are largepktrange consistently (at least 90%) very large (>1400 bytes) dnpercentsmall  90~100 Match when downstream packets are pktrange consistently (at least 90%) small packets (<200 bytes)

(76) A third rule shown in Table 4 defines the policy to be used to handle a VOIP (digital voice) call.

(77) TABLE-US-00004 TABLE 4 Parameter Value Explanation Name VoiceCall Provides the name for this rule. description “Traffic that is characteristic of Describes the type of traffic the rule a voice-only call (or is carrying and the desired policy for conference call) should be that kind of traffic. given wireless connectivity for low latency except when usage has been high. The first step down is to run just the downstream traffic is over satellite. With very high wireless usage both set of traffic should be carried satellite.” Score 200000 A relatively high score so that other rule matches will not take precedence should they occur. knobrange 1~6  Use this rule regardless of knob setting. upcontractid preferWireless, preferWireless, The policy is to use the use wireless preferWireless, preferWireless, except for the highest usage knob preferWireless, preferSatellite setting (knob level 6) dncontractid preferWireless, preferWireless, The policy is to use the wireless preferSatellite when wireless usage has been consistently low (usage knob levels 1 and 2) and to use satellite otherwise to conserve the more expensive wireless usage. multiperiodmatch 2~3  The rule matches when 2 of the last 3 measurement periods matched. upthruputrange 60~140 Match on moderately low upstream thruput dnthruputrange 60~120 Match on moderately low to moderate downstream thruput. dnppsrange 40~70  Match when 40 to 70 packets per second are being processed. upppsrange 40~70  Match when 40 to 70 packets per second are being processed. dnpercsmallinter 95~100 Match when the vast majority of arrivalpacketsrange packets have small interarrival times. uppercsmallinter 95~100 Match when the vast majority of arrivalpacketsrange packets have small interarrival times. dnmaxinterpacket 15~220 Match when the maximum inter- msrange arrival time is between 15 and 220 ms. upmaxinterpacket 15~220 Match when the maximum inter- msrange arrival time is between 15 and 220 ms. Dnpercentsmallp 92~100 Match when most packets are small ktrange (<200 bytes). Uppercentsmallp 92~100 Match when most packets are small ktrange (<200 bytes).

(78) A fourth example rule shown in Table 5 covers a conference call with downstream video higher-speed conference call with downstream video. This rule allows for more non-match samples. This rule may take around 10 seconds after the start of a call for the rule to become active and around 15 seconds after a call for it to end.

(79) TABLE-US-00005 TABLE 5 Parameter Value Explanation Name DnVideoConfCall Provides the name for this rule. description “Traffic that is characteristic Describes the type of traffic the rule of a downstream video is carrying and the desired policy for conference call with upstream that kind of traffic. voice traffic. This type of traffic should be given wireless connectivity for low latency except when usage has been high. The first step down is to run just the downstream traffic is over satellite. With very high wireless usage both set of traffic should be carried satellite.” Score 180000 A moderately high score so that other rule matches may or may not (depending on their score) take precedence should the match. knobrange 1~6  Use this rule regardless of knob setting. upcontractid preferWireless, The policy is to use the use wireless preferWireless, except for the highest usage knob preferWireless, setting (knob level 6) preferWireless, preferWireless, preferSatellite dncontractid preferWireless, preferSatellite The policy is to use the wireless when wireless usage has been consistently very low (usage knob levels 1) to minimize latency and to use satellite otherwise to conserve the more expensive wireless usage. multiperiodmatch 3~5  The rule matches when 3 of the last 5 measurement periods matched. upthruputrange 60~650 Match on moderately low upstream thruput but allow higher speeds as this has been empirically seen to occur. dnthruputrange 200~500  Match on moderate downstream thruput. dnppsrange 40~90  Match when 40 to 90 packets per second are being processed. upppsrange 40~90  Match when 40 to 90 packets per second are being processed. dnpercsmallinter 95~100 Match when the vast majority of arrivalpacketsrange packets have small interarrival times. uppercsmallinter 80~100 Match when the clear majority of arrivalpacketsrange packets have small interarrival times. dnmaxinterpacket 40~220 Match when the maximum inter- msrange arrival time is between 15 and 220 ms. upmaxinterpacket 40~500 Match when the maximum inter- msrange arrival time is between 15 and 220 ms. dnpercentsmall 60~100 Match when most packets are small pktrange (<200 bytes) but allow for large packets carrying video. uppercentsmall 80~100 Match when most packets are small pktrange (<200 bytes).

(80) Other rules may be utilized beyond the above examples to cover the various traffic types, including for example, the other traffic types discussed above. The rules may be empirically arrived at by examining the characteristics of the traffic and selecting a set of match criteria peculiar to that traffic.

(81) Usage Knob

(82) The usage knob (FIG. 2, 230, FIG. 3, 324) identifies how much LTE usage is to be reduced. The usage knob at any given time has a specific value that is referred to as its usage knob setting. The usage knob setting in a preferred embodiment govern LTE usage across a broad range of applications including web browsing, VOIP and video conference, but is, with the present invention, used to specifically govern a VPN connection's LTE usage.

(83) In the preferred embodiment the usage knob setting is based on the usage during the current and previous billing periods, where the usage knob setting increases based on any of a number of criteria including the following.

(84) Wireless usage billing period to date compared to a billing period limit.

(85) Wireless usage billing period to date as a function of a billing period limit and how much of the billing period has already taken place.

(86) Wireless usage billing period to date where the start of billing period usage is a function of how much usage occurred in the previous billing period.

(87) Many other such variations are possible in other embodiments but with the general principle that as the usage knob setting increases the level of LTE usage reduction should also increase.

(88) VPN Classifier Policy Determination

(89) The VPN Classifier, in the preferred embodiment, updates an outbound VPN IP flow's policy once every measurement period. The update is based on the current and recent measurement period measurements and only the Usage Knob setting.

(90) Candidate VPN Classification Rule Computation

(91) Typically, Usage Knob setting changes occurs infrequently. In the preferred embodiment, the VPN Classifier determines a candidate list of rules from the configured list of rules at startup and when the Usage Knob setting changes and when the configured list of rules change. The Candidate list of rules is the subset of configured rules which include the following. A knobrange which includes the current usage knob setting. Have a performance contract ID the current usage knob setting usable by the VPN's Classifier network element (upcontractid for the SD-WAN Router and dncontractid for the SD-WAN Gateway).

(92) High-Scoring VPN Classification Rule Determination

(93) In the preferred embodiment, at the end of each measurement period, the VPN Classifier updates the status of each candidate rule. It determines whether the period's measurements match the rule apart from the rule's multiperiodmatch parameter. With that determination it updates the rule's multiperiod long history (based on the N most recent periods) and determines whether the rule matches. The VPN Classifier then assigns the IP flow the SD-WAN selection policy of the highest-scoring rule whose status is “rule matches”. That policy is the highest scoring rule's contract ID for the current Usage Knob setting. The rules are organized with a default rule that will always match but with very low score and thus the VPN classifier will always have a highest scoring rule and thus be able to determine the current policy.

(94) Per-Packet Policy Assignment

(95) In the preferred embodiment, the VPN classifier, in addition to processing an outbound VPN packet to update its VPN Connection's statistics, tags the packet with its SD-WAN policy. It looks up its outbound IP flow and tags the packet with that flow's policy as determined at the end of the previous measurement period.

(96) FIG. 9 is a block diagram showing an example of router hardware 900 upon which aspects of this disclosure may be implemented. It will be understood that functional blocks illustrated in FIG. 9 are logical blocks of a hardware router such as a CPE router described in the functional block diagram of FIG. 2, and do not necessarily correspond to particular hardware. The router 900 may include a bus 902 or other communication mechanism for communicating information, and a processor 904 coupled with the bus 902 for processing information. The router 900 may also include a main memory 906, such as a random-access memory (RAM) or other dynamic storage device, coupled to the bus 902 for storing information and executable instructions to be executed by the processor 904. The executable instructions can include instruction that, when executed by the processor 904, cause the processor to perform operations in accordance with the flow diagrams and functionality described above. The main memory 906 may also be used for storing temporary variables or other intermediate information during execution of instructions to be executed by the processor 904.

(97) The router 900 may further include a read only memory (ROM) 908 or other static storage device coupled to the bus 902 for storing static information and instructions for the processor 904. A storage device 910, such as a flash or other non-volatile memory may be coupled to the bus 902 for storing information and instructions. The router 900 may also include a communication interface 912 coupled to a bus for two-way data communication coupling to a WAN 914 to provide data communication through one or more networks to other data devices. The router 900 may also include another communication interface 916 coupled to a bus for two-way data communication coupling to a LAN 918 to provide data communication through the LAN to other data devices such as LAN remote host 120 shown in FIG. 1.

(98) FIG. 10 is a block diagram showing an example a computer system 1000 upon which aspects of this disclosure may be implemented. It will be understood that functional blocks illustrated in FIG. 9 are logical blocks, and do not necessarily correspond to particular hardware.

(99) The computer system 1000 may include a bus 1002 or other communication mechanism for communicating information, and a processor 1004 coupled with the bus 1002 for processing information. The computer system 1000 may also include a main memory 1006, such as a random-access memory (RAM) or other dynamic storage device, coupled to the bus 1002 for storing information and executable instructions to be executed by the processor 1004. The executable instructions can include instruction that, when executed by the processor 1004, cause the processor to perform operations in accordance with the flow diagram of FIG. 8. The main memory 1006 may also be used for storing temporary variables or other intermediate information during execution of instructions to be executed by the processor 1004. The computer system 1000 may implement, for example, UT, SRAN, AMF, UPF, etc.

(100) The computer system 1000 may further include a read only memory (ROM) 1008 or other static storage device coupled to the bus 1002 for storing static information and instructions for the processor 1004. A storage device 1010, such as a flash or other non-volatile memory may be coupled to the bus 1002 for storing information and instructions.

(101) The computer system 1000 may be coupled via the bus 1002 to a display 1012, such as a liquid crystal display (LCD), for displaying information. One or more user input devices, such as the example user input device 1014 may be coupled to the bus 1002, and may be configured for receiving various user inputs, such as user command selections and communicating these to the processor 1004, or to the main memory 1006. The user input device 1014 may include physical structure, or virtual implementation, or both, providing user input modes or options, for controlling, for example, a cursor, visible to a user through display 1012 or through other techniques, and such modes or operations may include, for example virtual mouse, trackball, or cursor direction keys.

(102) The computer system 1000 may include respective resources of the processor 1004 executing, in an overlapping or interleaved manner, respective program instructions. Instructions may be read into the main memory 1006 from another machine-readable medium, such as the storage device 1010. In some examples, hard-wired circuitry may be used in place of or in combination with software instructions. The term “machine-readable medium” as used herein refers to any medium that participates in providing data that causes a machine to operate in a specific fashion. Such a medium may take forms, including but not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media may include, for example, optical or magnetic disks, such as storage device 1010. Transmission media may include optical paths, or electrical or acoustic signal propagation paths, and may include acoustic or light waves, such as those generated during radio-wave and infra-red data communications, that are capable of carrying instructions detectable by a physical mechanism for input to a machine.

(103) The computer system 1000 may also include a communication interface 1018 coupled to the bus 1002, for two-way data communication coupling to a network link 1020 connected to a local network 1022. The network link 1020 may provide data communication through one or more networks to other data devices. For example, the network link 1020 may provide a connection through the local network 1022 to a host computer 1024 or to data equipment operated by an Internet Service Provider (ISP) 1026 to access through the Internet 1028 a server 1030, for example, to obtain code for an application program.

(104) While various embodiments have been described, the description is intended to be exemplary, rather than limiting, and it is understood that many more embodiments and implementations are possible that are within the scope of the embodiments. Although many possible combinations of features are shown in the accompanying figures and discussed in this detailed description, many other combinations of the disclosed features are possible. Any feature of any embodiment may be used in combination with or substituted for any other feature or element in any other embodiment unless specifically restricted. Therefore, it will be understood that any of the features shown and/or discussed in the present disclosure may be implemented together in any suitable combination. Accordingly, the embodiments are not to be restricted except in light of the attached claims and their equivalents. Also, various modifications and changes may be made within the scope of the attached claims.

(105) While the foregoing has described what are considered to be the best mode and/or other examples, it is understood that various modifications may be made therein and that the subject matter disclosed herein may be implemented in various forms and examples, and that the teachings may be applied in numerous applications, only some of which have been described herein. It is intended by the following claims to claim any and all applications, modifications and variations that fall within the true scope of the present teachings.

(106) Unless otherwise stated, all measurements, values, ratings, positions, magnitudes, sizes, and other specifications that are set forth in this specification, including in the claims that follow, are approximate, not exact. They are intended to have a reasonable range that is consistent with the functions to which they relate and with what is customary in the art to which they pertain.

(107) The scope of protection is limited solely by the claims that now follow. That scope is intended and should be interpreted to be as broad as is consistent with the ordinary meaning of the language that is used in the claims when interpreted in light of this specification and the prosecution history that follows and to encompass all structural and functional equivalents. Notwithstanding, none of the claims are intended to embrace subject matter that fails to satisfy the requirement of Sections 101, 102, or 103 of the Patent Act, nor should they be interpreted in such a way. Any unintended embracement of such subject matter is hereby disclaimed.

(108) Except as stated immediately above, nothing that has been stated or illustrated is intended or should be interpreted to cause a dedication of any component, step, feature, object, benefit, advantage, or equivalent to the public, regardless of whether it is or is not recited in the claims.

(109) It will be understood that the terms and expressions used herein have the ordinary meaning as is accorded to such terms and expressions with respect to their corresponding respective areas of inquiry and study except where specific meanings have otherwise been set forth herein. Relational terms such as first and second and the like may be used solely to distinguish one entity or action from another without necessarily requiring or implying any actual such relationship or order between such entities or actions. The terms “comprises,” “comprising,” or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by “a” or “an” does not, without further constraints, preclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.