A live broadcast platform real-time interactive game system

Abstract

The application relates to the technical field of network live broadcast and real-time interactive game, and discloses a live broadcast platform real-time interactive game system, which comprises a game integration module, is used for providing an interactive game with a deterministic logic kernel and can export a game world state; a real-time matching and negotiation module, is used for matching a host and a game participating audience and negotiating and determining the capability parameters of each client; and a distributed live broadcast synchronization engine connected with the game integration module and the real-time matching and negotiation module, is used for coding the game world state into state vector data and broadcasting to multiple clients. The application adopts a difference vector synchronization and client local rendering scheme, replaces a traditional video stream, eliminates transmission delay, realizes extremely low delay synchronization, enables mass audiences to interact with the host in real time, and obtains a free and personalized watching experience.

Description

Technical Field

[0001] This invention relates to the field of online live streaming and real-time interactive gaming technology, specifically a real-time interactive gaming system for a live streaming platform. Background Technology

[0002] With the rapid development of internet technology, online live streaming has become an important way for people to obtain information and engage in entertainment. Traditional live streaming primarily involves the host unilaterally outputting video content, with viewers providing superficial feedback through comments or virtual gifts. However, this one-way, asynchronous interaction is gradually failing to meet users' growing demands for immersion and participation. Therefore, an interactive format combining real-time gaming with live streaming has emerged, offering new possibilities for enhancing user engagement and expanding business models.

[0003] In current technological practices, implementing such interactive games typically relies on encoding the streamer's gameplay footage and streaming it to the viewer's video player via a Content Delivery Network (CDN). This video streaming-based architecture suffers from a fundamental and insurmountable flaw: inherent high latency. From the moment the game footage is captured at the streamer's end, through encoding, network transmission, CDN distribution, and finally decoding and rendering at the viewer's end, the entire chain inevitably introduces a delay of several seconds or even longer. This latency means that the game footage seen by the viewer is severely lagging behind the actual gameplay progress of the streamer. Therefore, any interactive decisions made by the viewer are based on a "past" state, and their feedback cannot have an immediate and accurate impact on the current progress of the game, resulting in a significantly diminished interactive experience and weak engagement.

[0004] To alleviate latency issues, the industry has developed several low-latency live streaming protocols. However, these solutions essentially optimize video stream transmission efficiency and cannot reduce latency to the point of achieving frame-level synchronization. The fundamental obstacle to real-time interaction remains. On the other hand, while traditional multiplayer online games employ state synchronization technology to achieve low latency between players, these technologies are typically designed for a limited number of players with relatively equal network conditions. Applying them directly to live streaming scenarios with tens of thousands or even millions of viewers would incur prohibitive bandwidth costs due to broadcasting the complete game state to all clients. Furthermore, they cannot effectively address rendering consistency and smoothness issues arising from the vastly different performance levels of the viewers' devices.

[0005] Therefore, existing technologies have significant gaps in achieving large-scale, low-latency, and deeply integrated live interactive games. The market urgently needs a new technical architecture that can overcome the high latency constraints of video streaming and solve the applicability problems of traditional state synchronization technologies in ultra-large-scale concurrency and heterogeneous terminal scenarios, thereby providing broadcasters and all viewers with a truly unified, synchronized, and deeply engaging immersive interactive gaming experience. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a real-time interactive game system for live streaming platforms. It aims to solve the problem that existing live streaming technologies, due to their inherent high latency, high bandwidth costs, and shallow interaction modes, cannot achieve a truly synchronized and deeply integrated immersive game interaction experience between the streamer and a massive number of heterogeneous terminal viewers.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a real-time interactive game system for a live streaming platform, comprising: The game integration module is used to provide interactive games with a deterministic logic core and to export the game world state; The real-time matching and negotiation module is used to match streamers with viewers participating in the game and negotiate to determine the capability parameters of each client. A distributed live streaming synchronization engine, connected to the game integration module and the aforementioned real-time matching and negotiation module, is used to encode the game world state into state vector data and broadcast it to multiple clients; A unified interactive rendering interface, serving as a program for multiple clients, is used to receive the state vector data and perform real-time rendering on the client's local machine to present the game screen. The data statistics and analysis module is used to analyze system operation data.

[0008] Preferably, the distributed live streaming synchronization engine includes: The game state vectorization unit is used to encode the game world state exported from the game integration module into a layered state vector containing information from the core layer, presentation layer, and special effects layer.

[0009] Preferably, the distributed live streaming synchronization engine further includes: The predictive incremental coding unit is used to calculate the theoretical state of the current frame based on the authoritative state of the previous frame and the confirmed input instructions, using the same deterministic prediction function as the client. The theoretical state is then compared with the actual state of the current frame generated by the game state vectorization unit to generate the difference vector that contains only the difference between the two.

[0010] Preferably, the distributed live streaming synchronization engine further includes: The audience interaction aggregation unit is used to aggregate discrete audience interactions received from multiple unified interactive rendering interfaces into a continuous influence field that can produce physical effects in the game world coordinate system in real time, and inject the parameters of the influence field as part of the physical parameters into the core layer of the state vector. Among them, the field of influence at a certain point in time At world coordinates density Determined by the following formula: ; in, This indicates the summation of all interactions within the time window; Weights for interaction types; This is the kernel function used to smooth out influence. This is the projection point of the interaction in the world coordinate system; It is a time decay function; This is the timestamp of the interaction.

[0011] Preferably, the unified interactive rendering interface includes: The state synchronization and reconstruction unit is used to receive the difference vector. And through a structured merging operation, it is combined with the theoretical state generated by itself using the same deterministic prediction function. The process involves merging the data to accurately reconstruct the authoritative state of the current frame locally on the client. The reconstruction process follows the formula below: ; in, This represents a structured merging operation that overwrites the correction information contained in the difference vector onto the field corresponding to the theoretical state.

[0012] Preferably, the unified interactive rendering interface further includes: The layered real-time rendering unit is used to extract data of the corresponding layer from the reconstructed state vector according to the data subscription layer assigned to it by the real-time matching and negotiation module, and call the local graphics processing unit of the client for rendering.

[0013] Preferably, the real-time matching and negotiation module includes: The client capability negotiation unit is used to receive a list of device capability descriptions, including indicators such as graphics performance, memory, and bandwidth, submitted by the client when it accesses the network. And assign it an optimal state vector data subscription level. Wherein, the subscription level The determination is based on maximizing a comprehensive experience utility function. ; in, The subscription tier to be evaluated; This is used to evaluate the balance between visual appeal and smooth operation that can be achieved by subscribing to this level of data within the capabilities of the device.

[0014] Preferably, the real-time matching and negotiation module further includes: A real-time matching unit maintains a priority-based dynamic queue of applicants to select spectators to participate in the game; where the applicant's queuing weight... Determined using the following weighted summation model: ; in, For users' fan level or intimacy level; The value of virtual items consumed by users; This represents the user's cumulative waiting time. These are the configurable weight coefficients for each item.

[0015] Preferably, the state vector includes: The core layer contains the object transformation matrix, key event markers, and dynamic physical parameters generated by audience interaction, which are necessary to maintain the consistency of the game logic. The presentation layer contains non-critical animation state parameters used to enhance visual presentation. The special effects layer contains particle and lighting effect parameters that consume a lot of resources.

[0016] The present invention also provides a method for real-time interactive gaming on a live streaming platform, the method comprising the following steps: The game integration module provides and exports structured game world states; The real-time matching and negotiation module identifies the audience participating in the game and assigns data subscription strategies to all clients; The distributed live streaming synchronization engine encodes the game world state into difference vector data and sends it to multiple clients via broadcast. As a unified interactive rendering interface for multiple client programs, it receives the difference vector data and independently reconstructs the state and renders the screen in real time locally. The data statistics and analysis module analyzes the system's operational data.

[0017] This invention provides a real-time interactive game system for a live streaming platform. It has the following beneficial effects: 1. This invention employs a distributed live streaming synchronization engine to encode the game world state into difference vector data, and uses a unified interactive rendering interface to reconstruct the state and render it in real time on the client side. This completely eliminates the encoding, decoding, and transmission delays of traditional video streams, achieving extremely low latency end-to-end synchronization. This allows streamers, players, and a massive audience to be in the same latency-free game space, making true real-time interaction and personalized viewing possible.

[0018] 2. This invention introduces a predictive incremental coding unit, utilizing a deterministic prediction function shared by the server and client. Instead of broadcasting complete state data for each frame, it only needs to broadcast a highly sparse difference vector containing the prediction bias. This significantly reduces network bandwidth consumption, enabling the system to support a massive number of clients online simultaneously at extremely low cost. It effectively solves the technical bottleneck of traditional state synchronization schemes, which are difficult to commercially apply on a large scale due to excessively high bandwidth costs.

[0019] 3. This invention, by setting up an audience interaction aggregation unit, transforms the discrete interactive behaviors of a massive number of viewers into a continuous influence field that can produce physical effects in the game world in real time through a specific aggregation formula, and injects the parameters of this field into the core state vector. This design transforms viewers from passive "viewers" into "participants" who can directly influence the game process through collective will, greatly enhancing the sense of participation, immersion, and unpredictability of live streaming, creating entirely new interactive gameplay and business models, and significantly improving user stickiness.

[0020] 4. This invention, through a client-side capability negotiation unit within the real-time matching and negotiation module, dynamically assesses a user's device performance upon access and assigns the optimal data subscription level based on the experience utility function. The client's layered real-time rendering unit then renders the data on demand. This adaptive mechanism ensures a consistent and smooth core experience and differentiated visual effects across terminal devices with varying performance levels, effectively resolving the fragmented user experience caused by inconsistent device performance and significantly expanding the system's user reach.

[0021] 5. This invention, by adding a data statistics and analysis module, conducts in-depth analysis and mining of end-to-end system operational data, particularly the correlation between user interaction valence and key game events, terminal performance distribution, and network synchronization efficiency. This provides platform operators and game developers with objective and quantitative decision-making basis, enabling continuous optimization of network performance, balancing gameplay, and verification of the effectiveness of interaction mechanisms. It forms a data closed loop driving system self-iteration and content innovation, enhancing the platform's long-term operational capabilities and commercial value. Attached Figure Description

[0022] Figure 1This is a schematic diagram of the system architecture of the present invention; Figure 2 This is a schematic diagram of the game integration module architecture of the present invention; Figure 3 This is a schematic diagram of the real-time matching and negotiation module architecture of the present invention; Figure 4 This is a schematic diagram of the distributed live streaming synchronization engine architecture of the present invention; Figure 5 This is a schematic diagram of the unified interactive rendering interface architecture of the present invention; Figure 6 This is a schematic diagram of the data statistics and analysis module architecture of the present invention; Figure 7 This is a schematic diagram of the process flow of the method steps of the present invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see the appendix Figures 1-6 This invention provides a real-time interactive game system for a live streaming platform, comprising: The game integration module is used to provide interactive games with a deterministic logic core and to export the game world state; This embodiment provides a detailed description of the game integration module of the present invention. As the source of logic and content for the entire real-time interactive system, this module's core function is to provide an interactive game with a deterministic logical kernel and to export the structured game world state in real time according to the standard interface defined by this system. The stable and standardized implementation of this module is a fundamental prerequisite for the accurate and efficient execution of all subsequent distributed synchronization and rendering processes.

[0025] To clarify the game integration module described in this invention, it is first necessary to explain one of its core characteristics, namely, the deterministic logic kernel. In the technical context of this invention, determinism is the cornerstone of implementing a predictive synchronization mechanism. Its core essence lies in the fact that the evolution of the core gameplay and physical logic of any game integrated into this system must be completely reproducible. Specifically, given identical initial states (e.g., scene layout and character attributes at the start of a match) and a time-synchronized sequence of user input commands, the game engine, running in any independent hardware or software environment, must produce completely consistent state evolution results.

[0026] To achieve this determinism, the game integration module in this embodiment follows the following principles in its design: First, for all aspects of the game logic that require randomness, such as critical hit detection and random resource generation, pseudo-random number generators based on the local system clock or asynchronous seeds are not used. Preferably, the system uses a random seed uniformly distributed by the server at the start of the game to initialize a pseudo-random number generator (PRNG). All clients and servers use this PRNG and consume random numbers in the order of synchronized logical frames, thereby ensuring the consistency of random events across all terminals.

[0027] Second, in physical calculations and other processes involving floating-point operations, measures must be taken to avoid inconsistencies in calculation results caused by differences in the implementation of floating-point units across different hardware architectures (such as CPUs or GPUs). A preferred implementation is to use fixed-point algorithms instead of floating-point algorithms in core logic calculations. Alternatively, all participants can be forced to use a unified floating-point library and compilation options compliant with the IEEE 754 standard to maximize the consistency of cross-platform calculation results.

[0028] Third, for game engines that use multi-threading technology, it is necessary to strictly manage read and write operations on the game world state to avoid race conditions caused by the uncertainty of thread scheduling, and to ensure that the game state is unique and deterministic at the end of each logical frame.

[0029] Furthermore, another core function of this module is to implement a standardized StateVector Export Interface. This interface serves as a bridge between the complex, proprietary data structures within the game engine and the standardized state vectors used by the synchronization engine of this invention. At the end of each synchronization cycle (e.g., a logical frame), the game engine must use this interface to completely and losslessly convert its internal world's current "snapshot" into a hierarchical, high-dimensional set of state vectors. In this embodiment, the state vector The structure is predefined as a set containing multiple levels: ; The purpose of this layered design is to classify game states according to their importance and their consumption of rendering performance, thereby providing a data foundation for subsequent "client capability negotiation" and "layered real-time rendering".

[0030] Specifically, the core layer (Core Layer) contains the minimum dataset necessary to maintain absolute consistency of game logic across all clients. The game engine needs to populate the following information through an exported interface: ; in, It is a collection containing 4x4 world transformation matrices for all dynamic objects in the scene that are marked as "critical". These objects are typically player characters, key NPCs, important items, and other entities that directly affect gameplay. This matrix precisely defines their position, rotation, and scaling in the world coordinate system, and is the foundation for scene rendering and physics determination; It is a high-dimensional bit vector. Game developers predefine all possible instantaneous events in the game that require precise synchronization (e.g., character A deals damage to character B, character C uses a specific skill, team D scores, etc.) and assign a unique bit to each event. When an event occurs in the current frame, the game engine sets the corresponding bit in the vector to 1. This method is more efficient and accurate than transmitting complex event descriptions; It is a collection of key physical parameters that drive the evolution of the game world. This includes not only static parameters such as the global gravity constant and air drag coefficient, but more importantly, it also contains a reserved field to receive and apply the "influence field" parameters calculated and injected in real time by the "viewer interaction aggregation unit" in the downstream "distributed live streaming synchronization engine". In this way, the game engine's physics calculations can respond to the collective will of a massive audience.

[0031] Furthermore, the presentation layer The Presentation Layer primarily contains non-critical animation state data used to enhance visual appeal. Its typical structure is as follows: ; in, It is a collection containing detailed animation state machine parameters for each dynamic object, such as the index of the currently playing animation clip, the normalized playback time of the clip, and the weights of animation blending. Clients that subscribe to this layer of data can render smoother and more detailed character movements and scene representations.

[0032] Special effects layer This layer is specifically designed to handle state parameters for visual effects that demand high levels of graphics processing power from the client. Examples include the emitter state of a complex particle system (including its position, emission rate, particle lifecycle, color gradient, etc.) or the parameters of a dynamic light source requiring complex lighting calculations. Clients that do not subscribe to this layer will not render these resource-intensive effects, but their core game logic and experience will remain completely consistent with those that subscribe.

[0033] By implementing the aforementioned deterministic logic kernel and state vector export interface, the game integration module of this invention can stably and reliably transform any compliant game content into a standardized state data stream suitable for large-scale distributed synchronization. It not only provides high-quality "data raw materials" for the entire system's operation, but its inherent layered design philosophy also lays a solid technical foundation from the content source for achieving adaptive and personalized interactive experiences across massive heterogeneous terminals, thus ensuring the integrity and advancement of the technical solution of this invention.

[0034] The real-time matching and negotiation module is used to match streamers with viewers participating in the game and negotiate to determine the capability parameters of each client. This embodiment elaborates on the real-time matching and negotiation module of the present invention. This module plays a crucial role as both a "management entry point" and an "experience coordinator" in the overall system architecture. Its functions are not singular but multifaceted: firstly, it handles matching requests from a massive audience of users wishing to directly participate in the game, selecting players in an orderly and fair manner; secondly, it must negotiate and determine a data subscription strategy adapted to the capabilities of each connected client, whether they are players or spectators. This is a technical prerequisite for achieving a unified and smooth interactive experience across a large number of heterogeneous terminals.

[0035] To clarify the real-time matching function described in this invention, the Real-time Matching Unit within this module needs to be explained. When the streamer starts an interactive game, an application entry for participating in the game will appear on the unified interactive rendering interface on the viewer's end. The function of this unit is to receive and manage these applications. Preferably, this unit does not employ a simple first-come, first-served mechanism, but rather maintains a dynamic, priority-based queue of applicants. This design aims to balance the interests of multiple parties, providing streamers and the platform with more flexible operational methods.

[0036] In a preferred embodiment of the present invention, the user's queuing weight Quantification can be performed using a weighted summation model: ; in, For users' fan level or intimacy level; The value of virtual items consumed by users; This represents the user's cumulative waiting time. These are the configurable weight coefficients for each item.

[0037] When a new player needs to be selected, this unit selects the current player from the queue. The system identifies one or more users with the highest values ​​and passes their identity information (such as user ID) to the downstream distributed live streaming synchronization engine to establish their status as the authoritative source of game input.

[0038] The core innovation of this module lies in its Client Capability Negotiation Unit. The purpose of this unit is to fundamentally solve a core challenge faced by the distributed rendering architecture of this invention: how to ensure a logically synchronized, yet individually smooth viewing and interactive experience on a massive number of terminal devices with vastly different performance levels.

[0039] The workflow of this unit is as follows: When any user's client program (whether it is a selected player or a large number of spectators), namely the "unified interactive rendering interface" described in this invention, connects to the game session for the first time, it must first actively submit a structured device capability profile to this negotiation unit. This list summarizes the key performance indicators of the client's current operating environment.

[0040] The equipment capability description list It is a structured dataset, and its preferred composition can be represented as: ; in, A comprehensive performance benchmark score representing the client's graphics processing unit (GPU). Preferably, this score can be determined at startup using a lightweight benchmark program built into the client, or obtained by querying a pre-built performance database covering mainstream device models; This represents the current available memory capacity of the client device for this application, in MB. This represents the downlink network bandwidth estimated in real time by the client's network module, in Mbps. This parameter changes dynamically and reflects the user's current actual network conditions. This represents the physical screen resolution of the client device.

[0041] The negotiation unit on the server side received this list. Then, its core task is to assign an optimal state vector subscription level to the client. This decision directly relates to the granularity of data that the client will subsequently receive from the distributed live streaming synchronization engine. As mentioned earlier, the state vector... Designed to include a core layer , presentation layer and special effects layer The hierarchical structure.

[0042] This decision-making process can be mathematically modeled as a utility function maximization problem: ; in, It is the subscription tier that is ultimately determined for the client; For the subscription tier to be evaluated, This is used to evaluate the balance between visual appeal and smooth operation that can be achieved by subscribing to this level of data within the capabilities of the device.

[0043] After the decision is made, the negotiation unit will determine the subscription tier. The instruction is sent back to the corresponding client. Upon receiving the instruction, the client's unified interactive rendering interface requests the corresponding level of data stream from the distributed live streaming synchronization engine and configures its local rendering unit to ensure that the subsequent rendering process matches its device performance.

[0044] A distributed live streaming synchronization engine, connected to the game integration module and the aforementioned real-time matching and negotiation module, is used to encode the game world state into state vector data and broadcast it to multiple clients; This embodiment provides a detailed description of the distributed live streaming synchronization engine of the present invention. This engine is the hub and core of the technical solution of the present invention. Its upstream is connected to the game integration module and the real-time matching and negotiation module, while its downstream is a unified interactive rendering interface for all clients. The fundamental purpose of this engine is to completely abandon the traditional, high-latency video encoding and streaming technology paradigm, and instead adopt a completely new, state vector-based low-latency synchronization mechanism, thereby enabling true real-time interaction.

[0045] The engine's workflow is highly integrated and automated, and it can be further divided into multiple collaborative units, including: a game state vectorization unit, an audience interaction aggregation unit, and a predictive incremental coding unit.

[0046] First, the Game State Vectorization Unit (MSV) serves as the entry point for data processing, responsible for receiving and transforming the raw game world states exported from the upstream game integration module. Its core function is to translate the potentially proprietary data structures within the game engine into a set of predefined, standardized, and hierarchical state vectors defined by this system. This vectorization process forms the basis for all subsequent synchronization and compression operations. As mentioned earlier, this layered structure... The design aims to decouple different aspects of the game state in order to support downstream clients in adaptive layered rendering.

[0047] One particularly important detail to note is that this unit is involved in constructing the core layer state vector. physical parameter set in At that time, it doesn't simply replicate the static physics parameters of the game engine. It actively receives "influence field" parameters calculated in real time by the "audience interaction aggregation unit" within the engine. And it is seamlessly injected into as a dynamic variable. In this way, the collective will of a massive audience is quantified and solidified into a fundamental physical rule of the game world, which can then be consistently used by all clients for subsequent logical evolution calculations.

[0048] To achieve the aforementioned state injection, this engine includes a key innovative unit: the Audience Interaction Aggregation Unit. This unit was designed to address the disconnect between traditional interactive formats like "bullet comments" or "gifts" and the core gameplay. It aggregates the discrete and instantaneous interactions generated by a massive number of viewers on their respective clients into a continuous, dynamically changing macroscopic physical effect within the game world.

[0049] In a preferred embodiment of the present invention, the working process of this unit is as follows: First, it continuously receives standardized interaction vectors reported from various unified interactive rendering interfaces (clients). .in, It is a unique identifier for the user. These are the two-dimensional coordinates of the interactive behavior on the client's screen. It is the time when the interaction occurred, with a precise timestamp. This represents the type of interaction (e.g., regular click, paid click, etc.).

[0050] Next, the unit employs a field-theory-based aggregation model to process the tens of thousands of interaction vectors received instantaneously. These are aggregated in real time into a continuously distributed "audience influence field" within the three-dimensional game world. The shape and intensity of this field are determined by a dynamically changing density function. The mathematical expression described can be: ; in, Represents the three-dimensional coordinates of the game world At a point in time Influence density; This indicates the summation of all interactions within the time window; Weights for interaction types; This is the kernel function used to smooth out influence. This is the projection point of the interaction in the world coordinate system; It is a time decay function; The timestamp of the interaction; It is a time decay function, preferably an exponential decay function. Its function is to gradually weaken the influence of the interaction over time, ensuring the real-time and dynamic nature of the influence field.

[0051] The final calculated influence field parameters For example, it could be a density field. gradient field ( This parameter is used to simulate a "thrust" or "wind" field. It is then passed to the aforementioned game state vectorization unit to complete the injection.

[0052] At the end of the data processing pipeline of this engine is the Predictive Delta Encoding Unit. This unit is key to the invention's ability to support large-scale client synchronization with extremely low bandwidth consumption. Its purpose is to avoid directly broadcasting the massive complete state vector. Instead, it only broadcasts the changes over a very short period of time.

[0053] This unit employs a "prediction-correction" workflow: First, this unit internally holds the same deterministic state prediction function as all clients. At the beginning of each synchronization cycle, it utilizes the authority state of the previous frame. and the player input command set confirmed during this period Perform a local, non-rendering logical deduction to calculate a predicted state. .

[0054] At the same time, this unit obtains the real state containing the latest audience interaction impact from the game state vectorization unit. .

[0055] Next, the unit performs a structured comparison operation to generate a difference vector. This operation It's not a simple matrix subtraction, but a recursive comparison of two state vector tree structures, extracting only the nodes and values ​​that don't match. Because the predicted state... It already contains most of the deterministic evolution results, therefore the final difference vector Typically highly sparse, its content contains only parts that the prediction model cannot predict, such as deviations caused by new player actions during the period, new results generated by the deterministic random number generator, and unpredictable disturbances from the audience influence field.

[0056] Finally, the engine only uses this highly compressed difference vector It broadcasts to all clients connected to this game session via a preferred low-latency network transport protocol (such as QUIC).

[0057] A unified interactive rendering interface, serving as a program for multiple clients, is used to receive the state vector data and perform real-time rendering on the client's local machine to present the game screen. This embodiment provides a detailed description of the unified interactive rendering interface of the present invention. As an application running on multiple client devices, this interface plays a crucial "terminal executor" role in the technical solution of this invention. It fundamentally replaces the passive video player in traditional live streaming, transforming into an active, intelligent client program integrating state reconstruction, real-time rendering, and interactive generation. Its design and implementation are the final step in the present invention's transformation of abstract state data into a user-perceptible, personalized, and immersive interactive experience.

[0058] The internal functions of this unified interactive rendering interface can be undertaken by multiple logically decoupled but closely coordinated units, including: a state synchronization and reconstruction unit, a layered real-time rendering unit, and a user interaction generation unit.

[0059] First, the State Sync & Reconstruction Unit is the core of this interface for receiving and understanding server broadcast information. Its purpose is to perform the reverse operation of the "Predictive Incremental Encoding Unit" in the upstream "Distributed Live Streaming Synchronization Engine," thereby accurately restoring the complete and authoritative state of the game world locally on the client with extremely low data overhead.

[0060] In a preferred embodiment of the present invention, the workflow of this unit is a rigorous, two-stage process: The first stage is local state prediction. This unit integrates a deterministic state prediction function that is exactly the same as that on the server side. At the start of each synchronization frame, it is based on the authority state of the previous frame stored locally on the client. And the synchronized input command set of all players from the previous frame. This involves performing an independent, unrendered logical deduction. The output of this process is a local predicted state. Due to the function and input data It is completely consistent across all clients and servers, therefore, without considering new inputs and random events, all entities compute the same result. They are exactly the same.

[0061] The second stage is difference merging and state reconstruction. This unit receives a highly sparse difference vector broadcast from the server via the network module. Then, it performs a structured merge operation. The correction information contained in the difference vector is applied to the local predicted state in a "patch" manner. This allows for the reconstruction of the final authoritative state of the current frame. Its mathematical relationship can be expressed as: ; in, It is the final, authoritative game world state reconstructed on the client side, which is completely consistent with the current frame state of the server. It is the theoretical state of the current frame independently calculated by the client using a local prediction function; This represents a structured merge operation, which recursively traverses the structure of the difference vector and overwrites the corresponding fields of the local prediction state with the non-empty field values. It is a difference vector received from the server that contains only prediction bias and information about new events.

[0062] In this way, the client only needs to receive a very small amount of data to accurately reconstruct the complete game world locally, providing a data foundation for subsequent rendering.

[0063] Next, the Layered Real-time Rendering Unit is responsible for processing the abstract state vector reconstructed from the previous unit. This translates into vivid game visuals that are visible to the user. The core innovation of this unit lies in its "layered" and "adaptive" characteristics, and its behavior is directly controlled by the upstream "real-time matching and negotiation module".

[0064] Specifically, when the interface program starts up and connects to the game session, it receives a subscription tier assignment tailored to its needs from the negotiation module. This assignment is made by the server based on the client's list of device capabilities. This decision was made. In each subsequent rendering cycle, this rendering unit will strictly follow this assignment, from the complete state vector. In this process, data is selectively extracted and parsed.

[0065] For example, if a client is assigned Only the core layer is included. Then this unit will only parse and use The system uses object transformation matrices, events, and other information for basic rendering, while completely ignoring... and Animation and effects data are used in some cases. Conversely, a high-performance client may be allowed to use data at all levels to present the most complete visual effects. This on-demand data sourcing mechanism ensures that even devices with weaker performance can achieve a smooth rendering experience while maintaining the synchronization of core logic.

[0066] After extracting the data at the corresponding level, the unit calls the client device's local graphics processing unit (GPU) and graphics interface (preferably, modern graphics APIs such as Vulkan and Metal) to perform standard 3D scene rendering. Because the rendering process is entirely local, this invention allows all users, including onlookers, to freely adjust the camera's angle, distance, and orientation for their locally rendered video, obtaining a personalized, immersive viewing experience—a fundamental difference from passively receiving video streams from a fixed perspective.

[0067] Finally, the User Interaction Generation Unit (HEN) forms a bridge between the interface program and the user, responsible for capturing all the user's operational intentions and converting them into standardized data to be sent to the server, thereby closing the entire "observation-decision-action" interaction loop.

[0068] The unit's operating mode varies depending on the user's role. For users selected as players by the "Real-time Matchmaking and Negotiation Module," the unit captures all their game operation commands in a high-precision mode, such as keyboard keys, mouse clicks and movements, and gamepad joystick rotations. These commands are appended with precise client timestamps and immediately sent to the "Distributed Live Streaming Synchronization Engine" as authoritative input for calculating the game state of the next frame.

[0069] For a large audience, this section provides an interface specifically designed for large-scale interaction. When viewers click, swipe, or perform other actions on the rendered screen, this section converts these actions into a standardized interaction vector. Its data structure is as follows: ; in, It is a unique identifier for the user; The two-dimensional coordinates on the client's screen where the interaction occurs; It is a high-precision timestamp of the interaction occurrence given by the client's clock; It is an enumeration value representing the type of interaction, such as distinguishing between normal clicks and clicks for paid items.

[0070] These generated interaction vectors They will be sent in batches and efficiently to the "audience interaction aggregation unit" of the "distributed live streaming synchronization engine" to become the data source that constitutes the "audience influence field".

[0071] The data statistics and analysis module is used to analyze system operation data; This embodiment provides a detailed description of the data statistics and analysis module of the present invention. Within the overall architecture of the present invention, this module is positioned as an "offline brain" and "monitoring center" that operates as a bypass system, not in real-time. Its core function is not to process any real-time game logic or synchronization signaling, but rather to systematically collect, store, process, and deeply analyze the massive amounts of multi-dimensional data generated during the operation of the entire real-time interactive game system described in this invention. This aims to provide objective and quantitative data support for continuous system optimization, operational decision-making, and fault diagnosis.

[0072] It should be clarified that the "system operation data" analyzed by this module is a comprehensive dataset, whose sources cover all other core modules described in this invention, including but not limited to: the game integration module, the real-time matching and negotiation module, the distributed live streaming synchronization engine, and the unified interactive rendering interface as a client. This module collects this data through subscription or log retrieval, and it is then processed by multiple internal analysis units.

[0073] To clarify the data analysis function described in this invention, the terminal performance distribution analysis unit within this module needs to be explained. The main data source for this unit is the device capability description list collected by the "Real-time Matching and Negotiation Module" when processing client access. As mentioned above, the preferred structure of this list is: ; in, The overall performance benchmark score representing the client's graphics processing unit (GPU). This represents the current available memory capacity (MemorySize) of the client device for this application. This represents the downlink network bandwidth estimated in real time by the client's network module. This represents the physical screen resolution of the client device.

[0074] This unit will continuously and on a large scale aggregate all access users. Data. Through long-term statistical analysis of this data, the unit can create a dynamically updated panoramic picture of the terminal environment of the platform's active user group. For example, it can generate performance scores for the graphics processing unit. The distribution histogram clearly reveals the specific proportions of high-performance, medium-performance, and low-performance devices among platform users. Similarly, it can also analyze user network bandwidth. Geographic aggregation analysis was conducted to generate a heatmap of network quality. These analytical results provided the technical team with direct decision-making support regarding the performance level of devices to prioritize optimization in subsequent version iterations.

[0075] This module includes a synchronization performance and network analysis unit. This unit serves as a "barometer" of the overall system's core technology health. The data it analyzes primarily comes from the "distributed live streaming synchronization engine." The key performance indicators (KPIs) continuously monitored by this unit include: First, the difference vector The data size. This unit will calculate. Mean, median, and variance for different games and different time periods. Because The magnitude of the state prediction function directly reflects the state prediction function. The accuracy of the prediction model is generally considered good if it maintains a consistently low average level. Conversely, abnormal peaks or a continuously increasing trend may indicate that the prediction model is not covering certain scenarios in the game logic, requiring intervention from algorithm engineers for investigation and optimization.

[0076] Secondly, end-to-end synchronization latency. This unit calculates the end-to-end latency of state updates and statistically analyzes its distribution by analyzing the server's broadcast timestamp and the client's acknowledgment timestamp. This provides fundamental data for evaluating and optimizing network links and selecting network node deployment strategies. Thirdly, data compression ratio. This is achieved by comparing the broadcast difference vectors. With the uncompressed complete state vector Given the data size, this unit can quantitatively evaluate the actual performance of predictive incremental coding.

[0077] Furthermore, this module includes a highly innovative interaction valence correlation analysis unit. This unit aims to quantify the actual effectiveness of the unique "audience interaction aggregation" mechanism in this invention through data mining, specifically answering the core question, "To what extent does audience interaction truly influence the game?" To achieve this goal, this unit requires correlation analysis of information from multiple data sources.

[0078] In a preferred embodiment of the present invention, this unit constructs a correlation analysis model. The input variables of this model may include: Influence field parameters from the "distributed live streaming synchronization engine". For example, the audience influence field within a certain key time window. The peak intensity, its mathematical expression can be , where the density function .

[0079] Also from the engine are statistics about paid interactions, such as the total value or total number of paid interactions generated in a game.

[0080] The output variable (or target variable) of this model can be: Key event markers derived from game logic. For example, a Boolean variable indicating that "the tide of battle has turned." .

[0081] Macro-level metrics of a game, such as the total duration of the game. Or the user's day-one retention rate .

[0082] By employing methods such as statistical regression or machine learning classification, this unit can establish a mathematical relationship model between the aforementioned input and output variables. For example, a logistic regression model predicting the probability of a reversal in the battle situation: ; in, It is the Sigmoid function; These are the regression coefficients obtained by training the model using historical data; if the coefficients and If the statistical value is significantly positive, it provides strong data evidence for the argument that "audience interaction can substantially increase the entertainment value and unpredictability of the game." This is the output of the entire model, representing a probability value; This refers to the intercept or bias term of the model. and These are the regression coefficients of the model, which are the weight values ​​that the model learns by training on a large amount of historical data; This is an input feature representing the peak density of the "audience influence field" calculated by the audience interaction aggregation unit within the analysis time window. It is a quantitative indicator used to measure the intensity of collective audience interaction behavior at a given moment. This is another input feature, representing the total value of all paid interactions generated by the audience within the analysis's time window. It is a quantitative metric used to measure audience paid engagement. (The ellipsis indicates that the model is an scalable model. In addition to the two core features already listed, it can also include other input features that may affect the probability of a reversal in the game, such as game duration, player operation frequency, and the number of times specific game events occur, in order to further improve the model's prediction accuracy.

[0083] Please see Figure 7 The present invention also provides a distributed wearable health monitoring method, the method comprising the following steps: The game integration module provides and exports structured game world states; The real-time matching and negotiation module identifies the audience participating in the game and assigns data subscription strategies to all clients; The distributed live streaming synchronization engine encodes the game world state into difference vector data and sends it to multiple clients via broadcast. As a unified interactive rendering interface for multiple client programs, it receives the difference vector data and independently reconstructs the state and renders the screen in real time locally. The data statistics and analysis module analyzes the system's operational data. The method in this embodiment can be used to execute the above system embodiment, and its principle and technical effect are similar, so it will not be described again here.

[0084] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A real-time interactive game system for a live streaming platform, characterized in that, include: The game integration module is used to provide interactive games with a deterministic logic core and to export the game world state; The real-time matching and negotiation module is used to match streamers with viewers participating in the game and negotiate to determine the capability parameters of each client. A distributed live streaming synchronization engine, connected to the game integration module and the aforementioned real-time matching and negotiation module, is used to encode the game world state into state vector data and broadcast it to multiple clients; A unified interactive rendering interface, serving as a program for multiple clients, is used to receive the state vector data and perform real-time rendering on the client's local machine to present the game screen. The data statistics and analysis module is used to analyze system operation data.

2. The real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The distributed live streaming synchronization engine includes: The game state vectorization unit is used to encode the game world state exported from the game integration module into a layered state vector containing information from the core layer, presentation layer, and special effects layer.

3. The real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The distributed live streaming synchronization engine also includes: The predictive incremental coding unit is used to calculate the theoretical state of the current frame based on the authoritative state of the previous frame and the confirmed input instructions, using the same deterministic prediction function as the client. The theoretical state is then compared with the actual state of the current frame generated by the game state vectorization unit to generate the difference vector that contains only the difference between the two.

4. The real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The distributed live streaming synchronization engine also includes: The audience interaction aggregation unit is used to aggregate discrete audience interactions received from multiple unified interactive rendering interfaces into a continuous influence field that can produce physical effects in the game world coordinate system in real time, and inject the parameters of the influence field as part of the physical parameters into the core layer of the state vector. Among them, the field of influence at a certain point in time At world coordinates density Determined by the following formula: ； in, This indicates the summation of all interactions within the time window; Weights for interaction types; This is the kernel function used to smooth out influence. This is the projection point of the interaction in the world coordinate system; It is a time decay function; This is the timestamp of the interaction.

5. A real-time interactive game system for a live streaming platform according to claim 3, characterized in that, The unified interactive rendering interface includes: The state synchronization and reconstruction unit is used to receive the difference vector. And through a structured merging operation, it is combined with the theoretical state generated by itself using the same deterministic prediction function. The process involves merging the data to accurately reconstruct the authoritative state of the current frame locally on the client. The reconstruction process follows the formula below: ； in, This represents a structured merging operation that overwrites the correction information contained in the difference vector onto the field corresponding to the theoretical state.

6. A real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The unified interactive rendering interface also includes: The layered real-time rendering unit is used to extract data of the corresponding layer from the reconstructed state vector according to the data subscription layer assigned to it by the real-time matching and negotiation module, and call the local graphics processing unit of the client for rendering.

7. A real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The real-time matching and negotiation module includes: The client capability negotiation unit is used to receive a list of device capability descriptions, including indicators such as graphics performance, memory, and bandwidth, submitted by the client when it accesses the network. And assign it an optimal state vector data subscription level. Wherein, the subscription level The determination is based on maximizing a comprehensive experience utility function. ； in, The subscription tier to be evaluated; Used to evaluate the balance between visual appeal and smooth operation that can be achieved by subscribing to this level of data within the capabilities of the device.

8. A real-time interactive game system for a live streaming platform according to claim 1, characterized in that, The real-time matching and negotiation module also includes: A real-time matching unit maintains a priority-based dynamic queue of applicants to select spectators to participate in the game; where the applicant's queuing weight... Determined using the following weighted summation model: ； in, For users' fan level or intimacy level; The value of virtual items consumed by users; This represents the user's cumulative waiting time. These are the configurable weight coefficients for each item.

9. A real-time interactive game system for a live streaming platform according to claim 2, characterized in that, The state vector includes: The core layer contains the object transformation matrix, key event markers, and dynamic physical parameters generated by audience interaction, which are necessary to maintain the consistency of the game logic. The presentation layer contains non-critical animation state parameters used to enhance visual presentation. The special effects layer contains particle and lighting effect parameters that consume a lot of resources.

10. A method for real-time interactive gaming on a live streaming platform, characterized in that, For executing a real-time interactive game system on a live streaming platform as described in any one of claims 1-9, the system includes the following steps: The game integration module provides and exports structured game world states; The real-time matching and negotiation module identifies the audience participating in the game and assigns data subscription strategies to all clients; The distributed live streaming synchronization engine encodes the game world state into difference vector data and sends it to multiple clients via broadcast. As a unified interactive rendering interface for multiple client programs, it receives the difference vector data and independently reconstructs the state and renders the screen in real time locally. The data statistics and analysis module analyzes the system's operational data.

Images